difference between problem solving and creative thinking

Creative Thinking vs. Critical Thinking

What's the difference.

Creative thinking and critical thinking are two distinct but equally important cognitive processes. Creative thinking involves generating new ideas, concepts, and solutions by exploring various possibilities and thinking outside the box. It encourages imagination, originality, and innovation. On the other hand, critical thinking involves analyzing, evaluating, and questioning ideas, arguments, and information to make informed decisions and judgments. It emphasizes logical reasoning, evidence-based thinking, and the ability to identify biases and fallacies. While creative thinking focuses on generating ideas, critical thinking focuses on evaluating and refining those ideas. Both thinking processes are essential for problem-solving, decision-making, and personal growth.

Further Detail

Introduction.

Creative thinking and critical thinking are two distinct cognitive processes that play crucial roles in problem-solving, decision-making, and innovation. While they share some similarities, they also have distinct attributes that set them apart. In this article, we will explore the characteristics of creative thinking and critical thinking, highlighting their differences and showcasing how they complement each other in various contexts.

Creative Thinking

Creative thinking is a cognitive process that involves generating new ideas, concepts, or solutions by exploring possibilities, making connections, and thinking outside the box. It is characterized by originality, flexibility, and fluency of thought. Creative thinkers often challenge conventional wisdom, embrace ambiguity, and are open to taking risks. They are adept at finding alternative perspectives and exploring multiple solutions to problems.

One of the key attributes of creative thinking is the ability to think divergently. This means being able to generate a wide range of ideas or possibilities, often through brainstorming or free association. Creative thinkers are not limited by constraints and are willing to explore unconventional or unorthodox approaches to problem-solving.

Another important aspect of creative thinking is the ability to make connections between seemingly unrelated concepts or ideas. This skill, known as associative thinking, allows creative thinkers to draw upon a diverse range of knowledge and experiences to generate innovative solutions. They can see patterns, analogies, and relationships that others may overlook.

Furthermore, creative thinking involves the willingness to take risks and embrace failure as a learning opportunity. Creative thinkers understand that not all ideas will be successful, but they are not deterred by setbacks. They view failures as stepping stones towards finding the right solution and are persistent in their pursuit of innovative ideas.

In summary, creative thinking is characterized by divergent thinking, associative thinking, risk-taking, and persistence. It encourages the exploration of new ideas and unconventional approaches to problem-solving.

Critical Thinking

Critical thinking, on the other hand, is a cognitive process that involves analyzing, evaluating, and interpreting information to form reasoned judgments or decisions. It is characterized by logical, systematic, and objective thinking. Critical thinkers are skilled at identifying biases, assumptions, and fallacies in arguments, and they strive to make well-informed and rational decisions based on evidence.

One of the key attributes of critical thinking is the ability to think analytically. Critical thinkers break down complex problems or situations into smaller components, examine the relationships between them, and evaluate the evidence or information available. They are adept at identifying logical inconsistencies or flaws in reasoning, which helps them make sound judgments.

Another important aspect of critical thinking is the ability to evaluate information objectively. Critical thinkers are skeptical and question the validity and reliability of sources. They seek evidence, consider alternative viewpoints, and weigh the strengths and weaknesses of different arguments before forming their own opinions. This attribute is particularly valuable in today's information-rich society, where misinformation and biased narratives are prevalent.

Furthermore, critical thinking involves the ability to think systematically. Critical thinkers follow a logical and structured approach to problem-solving, ensuring that all relevant factors are considered. They are skilled at identifying assumptions, clarifying concepts, and drawing logical conclusions based on the available evidence. This systematic approach helps minimize errors and biases in decision-making.

In summary, critical thinking is characterized by analytical thinking, objective evaluation, skepticism, and systematic reasoning. It emphasizes the importance of evidence-based decision-making and helps individuals navigate complex and information-rich environments.

Complementary Attributes

While creative thinking and critical thinking have distinct attributes, they are not mutually exclusive. In fact, they often complement each other and can be seen as two sides of the same coin.

Creative thinking can benefit from critical thinking by providing a framework for evaluating and refining ideas. Critical thinking helps creative thinkers assess the feasibility, viability, and desirability of their innovative ideas. It allows them to identify potential flaws, consider alternative perspectives, and make informed decisions about which ideas to pursue further.

On the other hand, critical thinking can benefit from creative thinking by expanding the range of possibilities and solutions. Creative thinking encourages critical thinkers to explore unconventional approaches, challenge assumptions, and consider alternative viewpoints. It helps them break free from rigid thinking patterns and discover innovative solutions to complex problems.

Moreover, both creative thinking and critical thinking require open-mindedness and a willingness to embrace ambiguity. They both involve a certain level of discomfort and uncertainty, as individuals venture into uncharted territories of thought. By combining creative and critical thinking, individuals can develop a well-rounded cognitive toolkit that enables them to tackle a wide range of challenges.

Creative thinking and critical thinking are two distinct cognitive processes that bring unique attributes to problem-solving, decision-making, and innovation. Creative thinking emphasizes divergent thinking, associative thinking, risk-taking, and persistence, while critical thinking emphasizes analytical thinking, objective evaluation, skepticism, and systematic reasoning.

While they have their differences, creative thinking and critical thinking are not mutually exclusive. They complement each other and can be seen as two sides of the same coin. Creative thinking benefits from critical thinking by providing a framework for evaluation and refinement, while critical thinking benefits from creative thinking by expanding the range of possibilities and solutions.

By cultivating both creative and critical thinking skills, individuals can enhance their ability to navigate complex problems, make well-informed decisions, and drive innovation in various domains. These cognitive processes are not only valuable in academic and professional settings but also in everyday life, where the ability to think creatively and critically can lead to personal growth and success.

Comparisons may contain inaccurate information about people, places, or facts. Please report any issues.

Creative Thinking Definition

Creative thinking examples, why is creative thinking important, how to include creative thinking skills in a job application, how to build creativity, what is creative thinking definition and examples.

• Share on Twitter Share on Twitter
• Share on Facebook Share on Facebook
• Share on LinkedIn Share on LinkedIn

Forage puts students first. Our blog articles are written independently by our editorial team. They have not been paid for or sponsored by our partners. See our full  editorial guidelines .

Table of Contents

Creative thinking is the ability to come up with unique, original solutions. Also known as creative problem-solving, creative thinking is a valuable and marketable soft skill in a wide variety of careers. Here’s what you need to know about creative thinking at work and how to use it to land a job. 

Creative thinking is all about developing innovative solutions to problems. Creative thinkers brainstorm not only a large number of ideas but also a variety and range of them. In the workplace, creative thinking is highly valuable because employers look to hire innovative employees who can help them solve the company’s problems.

So, what does creative thinking in the workplace look like? First, a creative person brainstorms their ideas, then they’ll experiment with them. They look at ideas from multiple perspectives and examine how their solutions fit into the scope of what they’re working on. Creative thinkers aren’t afraid to take risks and try new ideas. In fact, this ability to develop, test, and implement original solutions makes them a valuable asset to just about any workplace. 

Creative thinking in the workplace might look like:

• Holding an interactive brainstorm to gather initial thoughts on a project
• Evaluating a current process and offering suggestions on how to improve it
• Researching other ways to market a product and leading experiments on new marketing channels
• Developing an innovative way to reach out to prospective clients
• Identifying a unique opportunity to promote the company brand and developing a strategy to do so
• Discovering a new way to measure a product initiative’s success and using learnings to iterate on the next version

Finding patterns in a company’s revenue growth and using data trends to strategize a new sales plan  

Creative thinking includes the process of innovative problem-solving — from analyzing the facts to brainstorming to working with others. Creative thinking examples include analytical skills, innovation, and collaboration.

Analytical Skills

Analytical skills are problem-solving skills that help you sort through facts, data, and information to develop rational solutions. These skills aid you in the first part of the creative thinking process as you brainstorm and start to generate ideas. 

Analytical skills include:

• Data analysis
• Forecasting
• Interpreting
• Communication

Innovation is the ability to come up with something new; however, you don’t need to develop the first flying car to be an innovative thinker. “Something new” at work might mean a method you haven’t tried before or experimenting with an unfamiliar process. Innovators in the workplace aren’t afraid to step away from tradition and explore something original, even if it might fail. 

Innovation skills include:

• Risk-taking
• Brainstorming
• Critical thinking

Collaboration

Creative thinking doesn’t have to happen alone; you might have your most creative ideas when bouncing your work off others. Collaboration skills ensure you consider multiple perspectives and ways of thinking when you develop and refine ideas.

Collaboration skills include:

• Written and verbal communication
• Active listening
• Inclusivity

A soft skill like creative thinking will always be valuable to employers, whether you’re looking for a marketing job or trying to land a career in finance . Employers need employees who can develop and experiment with new ideas to help them solve complex problems. 

“Many employers seek candidates that are analytical and outside-the-box thinkers which are iterations of creative thinking skills,” says Alejandra Garcia, manager, alumni college and career success at Code2College and Forage content development partner. “Thus, creative thinking, creative problem solving, innovative thinking, and analytical skills are all valuable in the current workplace — these skills are especially important in our ever-changing workplaces with new emerging technologies.”

The data supports this idea, too. According to the World Economic Forum’s 2023 Future of Jobs report , creative thinking is the second most important skill for workers in 2023, preceded only by analytical skills. Other top skills include soft skills like resilience, flexibility and agility, motivation and self-awareness, and curiosity and lifelong learning .

“The ability to navigate new challenges quickly can benefit any workplace!” Laura Fontenot, resume writing expert, ACRW, and CPRW, says. “The current world of work is fast-paced, technically driven, and constantly changing. Being intuitive, creative, driven, and a problem solver are key.”

If creative thinking is one of the top soft skills employers look for, how do you show you have it in a job application? The key is to prove these skills through examples of how you’ve used them rather than just naming them.

On a Resume

While creative thinking is a skill employers might look for, you don’t necessarily need to write “creative thinking” on your resume to show you have this skill. Instead, it’s better to demonstrate how you’ve used creative thinking skills to drive results.

“Think of your best mental strengths,” says Fontenot. “Are you a great problem solver? Do you understand how to phrase things differently? Can you learn a new skill quickly? Those questions can help you find great words for the resume . Consider adding things like problem-solving, intuition, collaboration, fast learner, organized, or communication.”

Log in to view and download a customizable resume template with examples of how to include creative thinking skills:

On Your Professional Profiles

You can show these skills outside of your resume in creative ways — including on your LinkedIn profile and website (if you have one!).

“Early professionals can make creative thinking a part of their professional brand by explicitly adding creative thinking or creative problem solving to their list of skills on their resumes and LinkedIn profiles — this will help with ATS optimizations,” Garcia advises. 

Yet beyond just listing this skill, Garcia adds that you can provide real proof of your creativity online, too.

“Consider adding projects or an online portfolio website link to your resume and LinkedIn where you can showcase projects you’ve worked on that demonstrate their problem-solving skills.”

In the Interview

In the interview , make sure you can describe your workflow and process for these projects or any other situation when you’ve used creative thinking. Elaborate how you brainstormed ideas, what range of ideas you had, how you tested and experimented, and how you decided on a final solution. 

It’s best to use the STAR method to structure your answers. This will ensure you clearly explain the situation and the results you brought by using your creative thinking skills.

>>MORE: Prepare to speak about your soft skills by practicing answers to commonly asked behavioral interview questions .

1. Put Yourself in a Box

Creative thinking is about “thinking outside the box,” but putting limitations on your problem-solving can help you think more freely and innovatively. For example, if someone tells you to make dinner, you may struggle to come up with a meal you don’t always cook. Yet if they ask you to make a hot dinner with three specific ingredients and two spices, you’ll more likely come up with something original. 

Putting yourself inside a box can help expand your thinking, whether that’s by telling yourself you need to include three charts in your presentation or giving yourself a strict word count for an article.

2. Switch up Your Routine

Routine can be a great productivity booster, but it also can get in the way of your creativity. So, switch up your routine for one project, day, or even an hour. This can be something as small as where you’re physically sitting when you do your work or something as big as your process for approaching projects. Challenging yourself to do something different will help you find creative ways to adapt to your new environment.

3. Challenge What’s Currently Working

Think about how you might expand or improve upon a current process. What would you do if you had more resources, whether that’s time, money, or another expert? What would you do if you had fewer resources? If this project was taking place at a different time of year? If the target audience was different? Imagining these different potential scenarios will force you to problem-solve and adjust for various (very possible!) circumstances. 

4. Find Inspiration

Creative thinking doesn’t happen in a bubble. It’s vital to ask for others’ opinions and ideas. Creative thinkers consider multiple perspectives and are curious about how others think. Ask your colleague about their work processes, whether it’s how they research for a client deliverable or how they approach meeting an external buyer. 

5. Ask for Feedback

The best way to improve a skill is to get feedback from others on how you’re using it — and you don’t need to set up a formal feedback session to do so. Instead, ask questions when you’re working with others about your work. Keep these questions open-ended and lead with curiosity instead of looking for a specific answer. What did they think of how you led the brainstorm? What would they have done differently? What strikes them about the final product? Keep an open mind and remember not to take the feedback personally. It’s an opportunity to grow, and growing those skills might just help you land your next job!

Two Sigma Professional Skills Development Program

Level up your non-technical skills and learn how to approach problems, set goals, and communicate clearly.

Avg. Time: 5-6 hours

Skills you’ll build: Project planning, project management, relationship management, explaining analysis

Image credit: Canva

Related Posts

6 negotiation skills to level up your work life, how to build conflict resolution skills: case studies and examples, what is github uses and getting started, upskill with forage.

Build career skills recruiters are looking for.

The science behind creativity

Psychologists and neuroscientists are exploring where creativity comes from and how to increase your own

Vol. 53 No. 3 Print version: page 40

• Neuropsychology
• Creativity and Innovation

Paul Seli, PhD, is falling asleep. As he nods off, a sleep-tracking glove called Dormio, developed by scientists at the Massachusetts Institute of Technology, detects his nascent sleep state and jars him awake. Pulled back from the brink, he jots down the artistic ideas that came to him during those semilucid moments.

Seli is an assistant professor of psychology and neuroscience at the Duke Institute for Brain Sciences and also an artist. He uses Dormio to tap into the world of hypnagogia, the transitional state that exists at the boundary between wakefulness and sleep. In a mini-experiment, he created a series of paintings inspired by ideas plucked from his hypnagogic state and another series from ideas that came to him during waking hours. Then he asked friends to rate how creative the paintings were, without telling them which were which. They judged the hypnagogic paintings as significantly more creative. “In dream states, we seem to be able to link things together that we normally wouldn’t connect,” Seli said. “It’s like there’s an artist in my brain that I get to know through hypnagogia.”

The experiment is one of many novel—and, yes, creative—ways that psychologists are studying the science of creativity. At an individual level, creativity can lead to personal fulfillment and positive academic and professional outcomes, and even be therapeutic. People take pleasure in creative thoughts, research suggests—even if they don’t think of themselves as especially creative. Beyond those individual benefits, creativity is an endeavor with implications for society, said Jonathan Schooler, PhD, a professor of psychological and brain sciences at the University of California, Santa Barbara. “Creativity is at the core of innovation. We rely on innovation for advancing humanity, as well as for pleasure and entertainment,” he said. “Creativity underlies so much of what humans value.”

In 1950, J. P. Guilford, PhD, then president of APA, laid out his vision for the psychological study of creativity ( American Psychologist , Vol. 5, No. 9, 1950). For half a century, researchers added to the scientific understanding of creativity incrementally, said John Kounios, PhD, an experimental psychologist who studies creativity and insight at Drexel University in Philadelphia. Much of that research focused on the personality traits linked to creativity and the cognitive aspects of the creative process.

But in the 21st century, the field has blossomed thanks to new advances in neuroimaging. “It’s become a tsunami of people studying creativity,” Kounios said. Psychologists and neuroscientists are uncovering new details about what it means to be creative and how to nurture that skill. “Creativity is of incredible real-world value,” Kounios said. “The ultimate goal is to figure out how to enhance it in a systematic way.”

Creativity in the brain

What, exactly, is creativity? The standard definition used by researchers characterizes creative ideas as those that are original and effective, as described by psychologist Mark A. Runco, PhD, director of creativity research and programming at Southern Oregon University ( Creativity Research Journal , Vol. 24, No. 1, 2012). But effectiveness, also called utility, is a slippery concept. Is a poem useful? What makes a sculpture effective? “Most researchers use some form of this definition, but most of us are also dissatisfied with it,” Kounios said.

Runco is working on an updated definition and has considered at least a dozen suggestions from colleagues for new components to consider. One frequently suggested feature is authenticity. “Creativity involves an honest expression,” he said.

Meanwhile, scientists are also struggling with the best way to measure the concept. As a marker of creativity, researchers often measure divergent thinking—the ability to generate a lot of possible solutions to a problem or question. The standard test of divergent thinking came from Guilford himself. Known as the alternate-uses test, the task asks participants to come up with novel uses for a common object such as a brick. But measures of divergent thinking haven’t been found to correlate well with real-world creativity. Does coming up with new uses for a brick imply a person will be good at abstract art or composing music or devising new methods for studying the brain? “It strikes me as using way too broad a brush,” Seli said. “I don’t think we measure creativity in the standard way that people think about creativity. As researchers, we need to be very clear about what we mean.”

One way to do that may be to move away from defining creativity based on a person’s creative output and focus instead on what’s going on in the brain, said Adam Green, PhD, a cognitive neuroscientist at Georgetown University and founder of the Society for the Neuroscience of Creativity . “The standard definition, that creativity is novel and useful, is a description of a product,” he noted. “By looking inward, we can see the process in action and start to identify the characteristics of creative thought. Neuroimaging is helping to shift the focus from creative product to creative process.”

That process seems to involve the coupling of disparate brain regions. Specifically, creativity often involves coordination between the cognitive control network, which is involved in executive functions such as planning and problem-solving, and the default mode network, which is most active during mind-wandering or daydreaming (Beaty, R. E., et al., Cerebral Cortex , Vol. 31, No. 10, 2021). The cooperation of those networks may be a unique feature of creativity, Green said. “These two systems are usually antagonistic. They rarely work together, but creativity seems to be one instance where they do.”

Green has also found evidence that an area called the frontopolar cortex, in the brain’s frontal lobes, is associated with creative thinking. And stimulating the area seems to boost creative abilities. He and his colleagues used transcranial direct current stimulation (tDCS) to stimulate the frontopolar cortex of participants as they tried to come up with novel analogies. Stimulating the area led participants to make analogies that were more semantically distant from one another—in other words, more creative ( Cerebral Cortex , Vol. 27, No. 4, 2017).

Green’s work suggests that targeting specific areas in the brain, either with neuromodulation or cognitive interventions, could enhance creativity. Yet no one is suggesting that a single brain region, or even a single neural network, is responsible for creative thought. “Creativity is not one system but many different mechanisms that, under ideal circumstances, work together in a seamless way,” Kounios said.

In search of the eureka moment

Creativity looks different from person to person. And even within one brain, there are different routes to a creative spark, Kounios explained. One involves what cognitive scientists call “System 1” (also called “Type 1”) processes: quick, unconscious thoughts—aha moments—that burst into consciousness. A second route involves “System 2” processes: thinking that is slow, deliberate, and conscious. “Creativity can use one or the other or a combination of the two,” he said. “You might use Type 1 thinking to generate ideas and Type 2 to critique and refine them.”

Which pathway a person uses might depend, in part, on their expertise. Kounios and his colleagues used electroencephalography (EEG) to examine what was happening in jazz musicians’ brains as they improvised on the piano. Then skilled jazz instructors rated those improvisations for creativity, and the researchers compared each musician’s most creative compositions. They found that for highly experienced musicians, the mechanisms used to generate creative ideas were largely automatic and unconscious, and they came from the left posterior part of the brain. Less-experienced pianists drew on more analytical, deliberative brain processes in the right frontal region to devise creative melodies, as Kounios and colleagues described in a special issue of NeuroImage on the neuroscience of creativity (Vol. 213, 2020). “It seems there are at least two pathways to get from where you are to a creative idea,” he said.

Coming up with an idea is only one part of the creative process. A painter needs to translate their vision to canvas. An inventor has to tinker with their concept to make a prototype that actually works. Still, the aha moment is an undeniably important component of the creative process. And science is beginning to illuminate those “lightbulb moments.”

Kounios examined the relationship between creative insight and the brain’s reward system by asking participants to solve anagrams in the lab. In people who were highly sensitive to rewards, a creative insight led to a burst of brain activity in the orbitofrontal cortex, the area of the brain that responds to basic pleasures like delicious food or addictive drugs ( NeuroImage , Vol. 214, 2020). That neural reward may explain, from an evolutionary standpoint, why humans seem driven to create, he said. “We seem wired to take pleasure in creative thoughts. There are neural rewards for thinking in a creative fashion, and that may be adaptive for our species.”

The rush you get from an aha moment might also signal that you’re onto something good, Schooler said. He and his colleagues studied these flashes of insight among creative writers and physicists. They surveyed the participants daily for two weeks, asking them to note their creative ideas and when they occurred. Participants reported that about a fifth of the most important ideas of the day happened when they were mind-wandering and not working on a task at hand ( Psychological Science , Vol. 30, No. 3, 2019). “These solutions were more likely to be associated with an aha moment and often overcoming an impasse of some sort,” Schooler said.

Six months later, the participants revisited those ideas and rated them for creative importance. This time, they rated their previous ideas as creative, but less important than they’d initially thought. That suggests that the spark of a eureka moment may not be a reliable clue that an idea has legs. “It seems like the aha experience may be a visceral marker of an important idea. But the aha experience can also inflate the meaningfulness of an idea that doesn’t have merit,” Schooler said. “We have to be careful of false ahas.”

Boosting your creativity

Much of the research in this realm has focused on creativity as a trait. Indeed, some people are naturally more creative than others. Creative individuals are more likely than others to possess the personality trait of openness. “Across different age groups, the best predictor of creativity is openness to new experiences,” said Anna Abraham, PhD, the E. Paul Torrance Professor and director of the Torrance Center for Creativity and Talent Development at the University of Georgia. “Creative people have the kind of curiosity that draws them toward learning new things and experiencing the world in new ways,” she said.

We can’t all be Thomas Edison or Maya Angelou. But creativity is also a state, and anyone can push themselves to be more creative. “Creativity is human capacity, and there’s always room for growth,” Runco said. A tolerant environment is often a necessary ingredient, he added. “Tolerant societies allow individuals to express themselves and explore new things. And as a parent or a teacher, you can model that creativity is valued and be open-minded when your child gives an answer you didn’t expect.”

One way to let your own creativity flow may be by tapping into your untethered mind. Seli is attempting to do so through his studies on hypnagogia. After pilot testing the idea on himself, he’s now working on a study that uses the sleep-tracking glove to explore creativity in a group of Duke undergrads. “In dream states, there seems to be connectivity between disparate ideas. You tend to link things together you normally wouldn’t, and this should lead to novel outcomes,” he said. “Neurally speaking, the idea is to increase connectivity between different areas of the brain.”

You don’t have to be asleep to forge those creative connections. Mind-wandering can also let the ideas flow. “Letting yourself daydream with a purpose, on a regular basis, might allow brain networks that don’t usually cooperate to literally form stronger connections,” Green said.

However, not all types of daydreams will get you there. Schooler found that people who engage in more personally meaningful daydreams (such as fantasizing about a future vacation or career change) report greater artistic achievement and more daily inspiration. People who are prone to fantastical daydreaming (such as inventing alternate realities or imaginary worlds) produced higher-quality creative writing in the lab and reported more daily creative behavior. But daydreams devoted to planning or problem-solving were not associated with creative behaviors ( Psychology of Aesthetics, Creativity, and the Arts , Vol. 15, No. 4, 2021).

It’s not just what you think about when you daydream, but where you are when you do it. Some research suggests spending time in nature can enhance creativity. That may be because of the natural world’s ability to restore attention, or perhaps it’s due to the tendency to let your mind wander when you’re in the great outdoors (Williams, K. J. H., et al., Journal of Environmental Psychology , Vol. 59, 2018). “A lot of creative figures go on walks in big, expansive environments. In a large space, your perceptual attention expands and your scope of thought also expands,” Kounios said. “That’s why working in a cubicle is bad for creativity. But working near a window can help.”

Wherever you choose to do it, fostering creativity requires time and effort. “People want the booster shot for creativity. But creativity isn’t something that comes magically. It’s a skill, and as with any new skill, the more you practice, the better you get,” Abraham said. In a not-yet-published study, she found three factors predicted peak originality in teenagers: openness to experience, intelligence, and, importantly, time spent engaged in creative hobbies. That is, taking the time to work on creative pursuits makes a difference. And the same is true for adults, she said. “Carve out time for yourself, figure out the conditions that are conducive to your creativity, and recognize that you need to keep pushing yourself. You won’t get to where you want to go if you don’t try.”

Those efforts can benefit your own sense of creative fulfillment and perhaps lead to rewards on an even grander scale. “I think everyday creativity is the most important kind,” Runco said. “If we can support the creativity of each and every individual, we’ll change the world.”

How to become more creative

1. Put in the work: People often think of creativity as a bolt of inspiration, like a lightbulb clicking on. But being creative in a particular domain—whether in the arts, in your work, or in your day-to-day life—is a skill. Carve out time to learn and practice.

2. Let your mind wander: Experts recommend “daydreaming with purpose.” Make opportunities to let your daydreams flow, while gently nudging them toward the creative challenge at hand. Some research suggests meditation may help people develop the habit of purposeful daydreaming.

3. Practice remote associations: Brainstorm ideas, jotting down whatever thoughts or notions come to you, no matter how wild. You can always edit later.

4. Go outside: Spending time in nature and wide-open spaces can expand your attention, enhance beneficial mind-wandering, and boost creativity.

5. Revisit your creative ideas: Aha moments can give you a high—but that rush might make you overestimate the merit of a creative idea. Don’t be afraid to revisit ideas to critique and tweak them later.

Further reading

Creativity: An introduction Kaufman, J. C., and Sternberg, R. J. (Eds.), Cambridge University Press, 2021

The eureka factor: Aha moments, creative insight, and the brain Kounios, J., & Beeman, M., Random House, 2015

Creativity anxiety: Evidence for anxiety that is specific to creative thinking, from STEM to the arts Daker, R. J., et al., Journal of Experimental Psychology: General , 2020

Predictors of creativity in young people: Using frequentist and Bayesian approaches in estimating the importance of individual and contextual factors Asquith, S. L., et al., Psychology of Aesthetics, Creativity, and the Arts , 2020

Recommended Reading

Contact apa, you may also like.

• Reviews / Why join our community?
• For companies
• Frequently asked questions

Creative Problem Solving

What is creative problem solving.

Creative problem solving (CPS) is a process that design teams use to generate ideas and solutions in their work. Designers and design teams apply an approach where they clarify a problem to understand it, ideate to generate good solutions, develop the most promising one, and implement it to create a successful solution for their brand’s users.  

© Creative Education Foundation, Fair Use

Why is Creative Problem Solving in UX Design Important?

Creative thinking and problem solving are core parts of user experience (UX) design. Note: the abbreviation “CPS” can also refer to cyber-physical systems. Creative problem solving might sound somewhat generic or broad. However, it’s an ideation approach that’s extremely useful across many industries.  

Not strictly a UX design-related approach, creative problem solving has its roots in psychology and education. Alex Osborn—who founded the Creative Education Foundation and devised brainstorming techniques—produced this approach to creative thinking in the 1940s. Along with Sid Parnes, he developed the Osborn-Parnes Creative Problem Solving Process. It was a new, systematic approach to problem solving and creativity fostering.  

Osborn’s CPS Process.

© IdeaSandbox.com, Fair Use

The main focus of the creative problem solving model is to improve creative thinking and generate novel solutions to problems. An important distinction exists between it and a UX design process such as design thinking. It’s that designers consider user needs in creative problem solving techniques, but they don’t necessarily have to make their users’ needs the primary focus. For example, a design team might trigger totally novel ideas from random stimuli—as opposed to working systematically from the initial stages of empathizing with their users. Even so, creative problem solving methods still tend to follow a process with structured stages. 

What are 4 Stages of Creative Problem Solving?

The model, adapted from Osborn’s original, typically features these steps:  

Clarify: Design teams first explore the area they want to find a solution within. They work to spot the challenge, problem or even goal they want to identify. They also start to collect data or information about it. It’s vital to understand the exact nature of the problem at this stage. So, design teams must build a clear picture of the issue they seek to tackle creatively. When they define the problem like this, they can start to question it with potential solutions.  

Ideate: Now that the team has a grasp of the problem that faces them, they can start to work to come up with potential solutions. They think divergently in brainstorming sessions and other ways to solve problems creatively, and approach the problem from as many angles as they can.  

Develop: Once the team has explored the potential solutions, they evaluate these and find the strongest and weakest qualities in each. Then, they commit to the one they decide is the best option for the problem at hand.  

Implement: Once the team has decided on the best fit for what they want to use, they discuss how to put this solution into action. They gauge its acceptability for stakeholders. Plus, they develop an accurate understanding of the activities and resources necessary to see it become a real, bankable solution.  

What Else does CPS Involve?

© Interaction Design Foundation, CC BY-SA 4.0

Two keys to the enterprise of creative problem solving are:  

Divergent Thinking

This is an ideation mode which designers leverage to widen their design space when they start to search for potential solutions. They generate as many new ideas as possible using various methods. For example, team members might use brainstorming or bad ideas to explore the vast area of possibilities. To think divergently means to go for:  

Quantity over quality: Teams generate ideas without fear of judgment (critically evaluating these ideas comes later). 

Novel ideas: Teams use disruptive and lateral thinking to break away from linear thinking and strive for truly original and extraordinary ideas.  

Choice creation: The freedom to explore the design space helps teams maximize their options, not only regarding potential solutions but also about how they understand the problem itself.  

Author and Human-Computer Interactivity Expert, Professor Alan Dix explains some techniques that are helpful for divergent thinking:  

• Transcript loading…

Convergent Thinking

This is the complementary half of the equation. In this ideation mode, designers analyze, filter, evaluate, clarify and modify the ideas they generated during divergent thinking. They use analytical, vertical and linear thinking to isolate novel and useful ideas, understand the design space possibilities and get nearer to potential solutions that will work best. The purpose with convergent thinking is to carefully and creatively:  

Look past logical norms (which people use in everyday critical thinking). 

Examine how an idea stands in relation to the problem.  

Understand the real dimensions of that problem.    

Professor Alan Dix explains convergent thinking in this video:  

What are the Benefits of Creative Problem Solving?

Design teams especially can benefit from this creative approach to problem solving because it:  

Empowers teams to arrive at a fine-grained definition of the problem they need to ideate over in a given situation.  

Gives a structured, learnable way to conduct problem-solving activities and direct them towards the most fruitful outcomes.  

Involves numerous techniques such as brainstorming and SCAMPER, so teams have more chances to explore the problem space more thoroughly.  

Can lead to large numbers of possible solutions thanks to a dedicated balance of divergent and convergent thinking.  

Values and nurtures designers and teams to create innovative design solutions in an accepting, respectful atmosphere.  

Is a collaborative approach that enables multiple participants to contribute—which makes for a positive environment with buy-in from those who participate.  

Enables teams to work out the most optimal solution available and examine all angles carefully before they put it into action.  

Is applicable in various contexts—such as business, arts and education—as well as in many areas of life in general.  

It’s especially crucial to see the value of creative problem solving in how it promotes out-of-the-box thinking as one of the valuable ingredients for teams to leverage.   

Watch as Professor Alan Dix explains how to think outside the box:  

How to Conduct Creative Problem Solving Best?

It’s important to point out that designers should consider—and stick to—some best practices when it comes to applying creative problem solving techniques. They should also adhere to some “house rules,” which the facilitator should define in no uncertain terms at the start of each session. So, designers and design teams should:  

Define the chief goal of the problem-solving activity: Everyone involved should be on the same page regarding their objective and what they want to achieve, why it’s essential to do it and how it aligns with the values of the brand. For example, SWOT analysis can help with this. Clarity is vital in this early stage.  Before team members can hope to work on ideating for potential solutions, they must recognize and clearly identify what the problem to tackle is.  

Have access to accurate information: A design team must be up to date with the realities that their brand faces, realities that their users and customers face, as well as what’s going on in the industry and facts about their competitors. A team must work to determine what the desired outcome is, as well as what the stakeholders’ needs and wants are. Another factor to consider in detail is what the benefits and risks of addressing a scenario or problem are—including the pros and cons that stakeholders and users would face if team members direct their attention on a particular area or problem.   

Suspend judgment: This is particularly important for two main reasons. For one, participants can challenge assumptions that might be blocking healthy ideation when they suggest ideas or elements of ideas that would otherwise seem of little value through a “traditional” lens. Second, if everyone’s free to suggest ideas without constraints, it promotes a calmer environment of acceptance—and so team members will be more likely to ideate better. Judgment will come later, in convergent thinking when the team works to tighten the net around the most effective solution. So, everyone should keep to positive language and encourage improvisational tactics—such as “yes…and”—so ideas can develop well.  

Balance divergent and convergent thinking: It’s important to know the difference between the two styles of thinking and when to practice them. This is why in a session like brainstorming, a facilitator must take control of proceedings and ensure the team engages in distinct divergent and convergent thinking sessions.  

Approach problems as questions: For example, “How Might We” questions can prompt team members to generate a great deal of ideas. That’s because they’re open-ended—as opposed to questions with “yes” or “no” answers. When a team frames a problem so freely, it permits them to explore far into the problem space so they can find the edges of the real matter at hand.  

UX Strategist and Consultant, William Hudson explains “How Might We” questions in this video:  

Use a variety of ideation methods: For example, in the divergent stage, teams can apply methods such as random metaphors or bad ideas to venture into a vast expanse of uncharted territory. With random metaphors, a team prompts innovation by drawing creative associations. With bad ideas, the point is to come up with ideas that are weird, wild and outrageous, as team members can then determine if valuable points exist in the idea—or a “bad” idea might even expose flaws in conventional ways of seeing problems and situations.  

Professor Alan Dix explains important points about bad ideas:  

• Copyright holder: William Heath Robinson. Appearance time: 1:30 - 1:33 Copyright license and terms: Public domain. Link: https://upload.wikimedia.org/wikipedia/commons/c/c9/William_Heath_Robinson_Inventions_-_Page_142.png
• Copyright holder: Rev Stan. Appearance time: 1:40 - 1:44 Copyright license and terms: CC BY 2.0 Link: As yummy as chocolate teapot courtesy of Choccywoccydoodah… _ Flickr.html
• Copyright holder: Fabel. Appearance time: 7:18 - 7:24 Copyright license and terms: CC BY-SA 3.0 Link: https://commons.wikimedia.org/wiki/File:Hammer_nails_smithonian.jpg
• Copyright holder: Marcus Hansson. Appearance time: 05:54 - 05:58 Copyright license and terms: CC BY 2.0 Link: https://www.flickr.com/photos/marcus_hansson/7758775386

What Special Considerations Should Designers Have for CPS?

Creative problem solving isn’t the only process design teams consider when thinking of potential risks. Teams that involve themselves in ideation sessions can run into problems, especially if they aren’t aware of them. Here are the main areas to watch:  

Bias is natural and human. Unfortunately, it can get in the way of user research and prevent a team from being truly creative and innovative. What’s more, it can utterly hinder the iterative process that should drive creative ideas to the best destinations. Bias takes many forms. It can rear its head without a design team member even realizing it. So, it’s vital to remember this and check it. One team member may examine an angle of the problem at hand and unconsciously view it through a lens. Then, they might voice a suggestion without realizing how they might have framed it for team members to hear. Another risk is that other team members might, for example, apply confirmation bias and overlook important points about potential solutions because they’re not in line with what they’re looking for.  

Professor Alan Dix explains bias and fixation as obstacles in creative problem solving examples, and how to overcome them:  

Conventionalism

Even in the most hopeful ideation sessions, there’s the risk that some team members may slide back to conventional ways to address a problem. They might climb back inside “the box” and not even realize it. That’s why it’s important to mindfully explore new idea territories around the situation under scrutiny and not merely toy with the notion while clinging to a default “traditional” approach, just because it’s the way the brand or others have “always done things.”   

Dominant Personalities and Rank Pulling

As with any group discussion, it’s vital for the facilitator to ensure that everyone has the chance to contribute. Team members with “louder” personalities can dominate the discussions and keep quieter members from offering their thoughts. Plus, without a level playing field, it can be hard for more junior members to join in without feeling a sense of talking out of place or even a fear of reprisal for disagreeing with senior members.  

Another point is that ideation sessions naturally involve asking many questions, which can bring on two issues. First, some individuals may over-defend their ideas as they’re protective of them. Second, team members may feel self-conscious as they might think if they ask many questions that it makes them appear frivolous or unintelligent. So, it’s vital for facilitators to ensure that all team members can speak up and ask away, both in divergent thinking sessions when they can offer ideas and convergent thinking sessions when they analyze others’ ideas.  

Premature Commitment

Another potential risk to any creativity exercise is that once a team senses a solution is the “best” one, everyone can start to shut off and overlook the chance that an alternative may still arise. This could be a symptom of ideation fatigue or a false consensus that a proposed solution is infallible. So, it’s vital that team members keep open minds and try to catch potential issues with the best-looking solution as early as possible. The key is an understanding of the need for iteration—something that’s integral to the design thinking process, for example.   

Overall, creative problem solving can help give a design team the altitude—and attitude—they need to explore the problem and solution spaces thoroughly. Team members can leverage a range of techniques to trawl through the hordes of possibilities that exist for virtually any design scenario. As with any method or tool, though, it takes mindful application and awareness of potential hazards to wield it properly. The most effective creative problem-solving sessions will be ones that keep “creative,” “problem” and “solving” in sharp focus until what emerges for the target audience proves to be more than the sum of these parts.  

Learn More About Creative Problem Solving

Take our course, Creativity: Methods to Design Better Products and Services . 

Watch our Master Class Harness Your Creativity To Design Better Products with Alan Dix, Professor, Author and Creativity Expert. 

Read our piece, 10 Simple Ideas to Get Your Creative Juices Flowing . 

Go to Exploring the Art of Innovation: Design Thinking vs. Creative Problem Solving by Marcino Waas for further details. 

Consult Creative Problem Solving by Harrison Stamell for more insights.  

Read The Osborn Parnes Creative Problem-Solving Process by Leigh Espy for additional information.  

See History of the creative problem-solving process by Jo North for more on the history of Creative Problem Solving. 

Questions about Creative Problem Solving

To start with, work to understand the user’s needs and pain points. Do your user research—interviews, surveys and observations are helpful, for instance. Analyze this data so you can spot patterns and insights. Define the problem clearly—and it needs to be extremely clear for the solution to be able to address it—and make sure it lines up with the users’ goals and your project’s objectives. 

You and your design team might hold a brainstorming session. It could be a variation such as brainwalking—where you move about the room ideating—or brainwriting, where you write down ideas. Alternatively, you could try generating weird and wonderful notions in a bad ideas ideation session. 

There’s a wealth of techniques you can use. In any case, engage stakeholders in brainstorming sessions to bring different perspectives on board the team’s trains of thought. What’s more, you can use tools like a Problem Statement Template to articulate the problem concisely. 

Take our course, Creativity: Methods to Design Better Products and Services . 

Watch as Author and Human-Computer Interaction Expert, Professor Alan Dix explains important points about bad ideas:  

Some things you might try are:  1. Change your environment: A new setting can stimulate fresh ideas. So, take a walk, visit a different room, or work outside. 

2. Try to break the problem down into smaller parts: Focus on just one piece at a time—that should make the task far less overwhelming. Use techniques like mind mapping so you can start to visualize connections and come up with ideas. 

3. Step away from work and indulge in activities that relax your mind: Is it listening to music for you? Or how about drawing? Or exercising? Whatever it is, if you break out of your routine and get into a relaxation groove, it can spark new thoughts and perspectives. 

4. Collaborate with others: Discuss the problem with colleagues, stakeholders, or—as long as you don’t divulge sensitive information or company secrets—friends. It can help you to get different viewpoints, and sometimes those new angles and fresh perspectives can help unlock a solution. 

5. Set aside dedicated time for creative thinking: Take time to get intense with creativity; prevent distractions and just immerse yourself in the problem as fully as you can with your team. Use techniques like brainstorming or the "Six Thinking Hats" to travel around the problem space and explore a wealth of angles. 

Remember, a persistent spirit and an open mind are key; so, keep experimenting with different approaches until you get that breakthrough. 

Watch as Professor Alan Dix explains important aspects of creativity and how to handle creative blocks: 

Read our piece, 10 Simple Ideas to Get Your Creative Juices Flowing . 

Watch as Professor Alan Dix explains the Six Thinking Hats ideation technique. 

Creative thinking is about coming up with new and innovative ideas by looking at problems from different angles—and imagining solutions that are truly fresh and unique. It takes an emphasis on divergent thinking to get “out there” and be original in the problem space. You can use techniques like brainstorming, mind mapping and free association to explore hordes of possibilities, many of which might be “hiding” in obscure corners of your—or someone on your team’s—imagination. 

Critical thinking is at the other end of the scale. It’s the convergent half of the divergent-convergent thinking approach. In that approach, once the ideation team have hauled in a good catch of ideas, it’s time for team members to analyze and evaluate these ideas to see how valid and effective each is. Everyone strives to consider the evidence, draw logical connections and eliminate any biases that could be creeping in to cloud judgments. Accuracy, sifting and refining are watchwords here. 

Watch as Professor Alan Dix explains divergent and convergent thinking: 

The tools you can use are in no short supply, and they’re readily available and inexpensive, too. Here are a few examples: 

Tools like mind maps are great ways to help you visualize ideas and make connections between them and elements within them. Try sketching out your thoughts and see how they relate to each other—you might discover unexpected gems, or germs of an idea that can splinter into something better, with more thought and development. 

The SCAMPER technique is another one you can try. It can help you catapult your mind into a new idea space as you Substitute, Combine, Adapt, Modify, Put to another use, Eliminate, and Reverse aspects of the problem you’re considering. 

The “5 Whys” technique is a good one to drill down to root causes with. Once you’ve spotted a problem, you can start working your way back to see what’s behind it. Then you do the same to work back to the cause of the cause. Keep going; usually five times will be enough to see what started the other problems as the root cause. 

Watch as the Father of UX Design, Don Norman explains the 5 Whys technique: 

Read all about SCAMPER in our topic definition of it. 

It’s natural for some things to get in the way of being creative in the face of a problem. It can be challenging enough to ideate creatively on your own, but it’s especially the case in group settings. Here are some common obstacles: 

1. Fear of failure or appearing “silly”: when people worry about making mistakes or sounding silly, they avoid taking risks and exploring new ideas. This fear stifles creativity. That’s why ideation sessions like bad ideas are so valuable—it turns this fear on its head. 

2. Rigid thinking: This can also raise itself as a high and thick barrier. If someone in an ideation session clings to established ways to approach problems (and potential solutions), it can hamper their ability to see different perspectives, let alone agree with them. They might even comment critically to dampen what might just be the brightest way forward. It takes an open mind and an awareness of one’s own bias to overcome this. 

3. Time pressure and resource scarcity: When a team has tight deadlines to work to, they may rush to the first workable solution and ignore a wide range of possibilities where the true best solution might be hiding. That’s why stakeholders and managers should give everyone enough time—as well as any needed tools, materials and support—to ideate and experiment. The best solution is in everybody’s interest, after all.  

It takes a few ingredients to get the environment just right for creative problem solving:  

Get in the mood for creativity: This could be a relaxing activity before you start your session, or a warm-up activity in the room. Then, later, encourage short breaks—they can rejuvenate the mind and help bring on fresh insights.  

Get the physical environment just right for creating problem solving: You and your team will want a comfortable and flexible workspace—preferably away from your workstations. Make sure the room is one where people can collaborate easily and also where they can work quietly. A meeting room is good as it will typically have room for whiteboards and comfortable space for group discussion. Note: you’ll also need sticky notes and other art supplies like markers. 

Make the atmosphere conducive for creative problem solving: Someone will need to play facilitator so everyone has some ground rules to work with. Encourage everyone to share ideas, that all ideas are valuable, and that egos and seniority have no place in the room. Of course, this may take some enforcement and repetition—especially as "louder" team members may try to dominate proceedings, anyway, and others may be self-conscious about sounding "ridiculous." 

Make sure you’ve got a diverse team: Diversity means different perspectives, which means richer and more innovative solutions can turn up. So, try to include individuals with different backgrounds, skills and viewpoints—sometimes, non-technical mindsets can spot ideas and points in a technical realm, which experienced programmers might miss, for instance. 

Watch our Master Class Harness Your Creativity To Design Better Products with Alan Dix, Professor, Author and Creativity Expert. 

Ideating alone? Watch as Professor Alan Dix gives valuable tips about how to nurture creativity: 

• Copyright holder: GerritR. Appearance time: 6:54 - 6:59 Copyright license and terms: CC-BY-SA-4.0 Link: https://commons.wikimedia.org/wiki/File:Blick_auf_das_Dylan_Thomas_Boathouse_und_die_Trichterm%C3%BCndung_des_Taf,_Wales.jpg

Research plays a crucial role in any kind of creative problem solving, and in creative problem solving itself it’s about collecting information about the problem—and, by association, the users themselves. You and your team members need to have a well-defined grasp of what you’re facing before you can start reaching out into the wide expanses of the idea space.  

Research helps you lay down a foundation of knowledge and avoid reinventing the wheel. Also, if you study existing solutions and industry trends, you’ll be able to understand what has worked before and what hasn't.  

What’s more, research is what will validate the ideas that come out of your ideation efforts. From testing concepts and prototypes with real users, you’ll get precious input about your creative solutions so you can fine-tune them to be innovative and practical—and give users what they want in a way that’s fresh and successful. 

Watch as UX Strategist and Consultant, William Hudson explains important points about user research: 

First, it’s crucial for a facilitator to make sure the divergent stage of the creative problem solving is over and your team is on to the convergent stage. Only then should any analysis happen.  

If others are being critical of your creative solutions, listen carefully and stay open-minded. Look on it as a chance to improve, and don’t take it personally. Indeed, the session facilitator should moderate to make sure everyone understands the nature of constructive criticism.  

If something’s unclear, be sure to ask the team member to be more specific, so you can understand their points clearly. 

Then, reflect on what you’ve heard. Is it valid? Something you can improve or explain? For example, in a bad ideas session, there may be an aspect of your idea that you can develop among the “bad” parts surrounding it. 

So, if you can, clarify any misunderstandings and explain your thought process. Just stay positive and calm and explain things to your critic and other team member. The insights you’ve picked up may strengthen your solution and help to refine it. 

Last—but not least—make sure you hear multiple perspectives. When you hear from different team members, chances are you’ll get a balanced view. It can also help you spot common themes and actionable improvements you might make. 

Watch as Todd Zaki Warfel, Author, Speaker and Leadership Coach, explains how to present design ideas to clients, a valuable skill in light of discussing feedback from stakeholders. 

Lateral thinking is a technique where you approach problems from new and unexpected angles. It encourages you to put aside conventional step-by-step logic and get “out there” to explore creative and unorthodox solutions. Author, physician and commentator Edward de Bono developed lateral thinking as a way to help break free from traditional patterns of thought. 

In creative problem solving, you can use lateral thinking to come up with truly innovative ideas—ones that standard logical processes might overlook. It’s about bypassing these so you can challenge assumptions and explore alternatives that point you and your team to breakthrough solutions. 

You can use techniques like brainstorming to apply lateral thinking and access ideas that are truly “outside the box” and what your team, your brand and your target audience really need to work on. 

Professor Alan Dix explains lateral thinking in this video: 

1. Baer, J. (2012). Domain Specificity and The Limits of Creativity Theory . The Journal of Creative Behavior, 46(1), 16–29.   John Baer's influential paper challenged the notion of a domain-general theory of creativity and argued for the importance of considering domain-specific factors in creative problem solving. This work has been highly influential in shaping the understanding of creativity as a domain-specific phenomenon and has implications for the assessment and development of creativity in various domains. 

2. Runco, M. A., & Jaeger, G. J. (2012). The Standard Definition of Creativity . Creativity Research Journal, 24(1), 92–96.   Mark A. Runco and Gerard J. Jaeger's paper proposed a standard definition of creativity, which has been widely adopted in the field. They defined creativity as the production of original and effective ideas, products, or solutions that are appropriate to the task at hand. This definition has been influential in providing a common framework for creativity research and assessment. 

1. Fogler, H. S., LeBlanc, S. E., & Rizzo, B. (2014). Strategies for Creative Problem Solving (3rd ed.). Prentice Hall. 

This book focuses on developing creative problem-solving strategies, particularly in engineering and technical contexts. It introduces various heuristic problem-solving techniques, optimization methods, and design thinking principles. The authors provide a systematic framework for approaching ill-defined problems, generating and implementing solutions, and evaluating the outcomes. With its practical exercises and real-world examples, this book has been influential in equipping professionals and students with the skills to tackle complex challenges creatively. 

2. De Bono, E. (1985). Six Thinking Hats . Little, Brown and Company.   

Edward de Bono's Six Thinking Hats introduces a powerful technique for parallel thinking and decision-making. The book outlines six different "hats" or perspectives that individuals can adopt to approach a problem or situation from various angles. This structured approach encourages creative problem-solving by separating different modes of thinking, such as emotional, logical, and creative perspectives. De Bono's work has been highly influential in promoting lateral thinking and providing a practical framework for group problem solving. 

3. Osborn, A. F. (1963). Applied Imagination: Principles and Procedures of Creative Problem-Solving (3rd ed.). Charles Scribner's Sons.  

Alex F. Osborn's Applied Imagination is a pioneering work that introduced the concept of brainstorming and other creative problem-solving techniques. Osborn emphasized how important it is to defer judgment and generate a large quantity of ideas before evaluating them. This book laid the groundwork for many subsequent developments in the field of creative problem-solving, and it’s been influential in promoting the use of structured ideation processes in various domains. 

Answer a Short Quiz to Earn a Gift

What is the first stage in the creative problem-solving process?

• Implementation
• Idea Generation
• Problem Identification

Which technique is commonly used during the idea generation stage of creative problem-solving?

• Brainstorming
• Prototyping

What is the main purpose of the evaluation stage in creative problem-solving?

• To generate as many ideas as possible
• To implement the solution
• To assess the feasibility and effectiveness of ideas

In the creative problem-solving process, what often follows after implementing a solution?

• Testing and Refinement

Which stage in the creative problem-solving process focuses on generating multiple possible solutions?

Better luck next time!

Do you want to improve your UX / UI Design skills? Join us now

Congratulations! You did amazing

You earned your gift with a perfect score! Let us send it to you.

Check Your Inbox

We’ve emailed your gift to [email protected] .

Literature on Creative Problem Solving

Here’s the entire UX literature on Creative Problem Solving by the Interaction Design Foundation, collated in one place:

Learn more about Creative Problem Solving

Take a deep dive into Creative Problem Solving with our course Creativity: Methods to Design Better Products and Services .

The overall goal of this course is to help you design better products, services and experiences by helping you and your team develop innovative and useful solutions. You’ll learn a human-focused, creative design process.

We’re going to show you what creativity is as well as a wealth of ideation methods ―both for generating new ideas and for developing your ideas further. You’ll learn skills and step-by-step methods you can use throughout the entire creative process. We’ll supply you with lots of templates and guides so by the end of the course you’ll have lots of hands-on methods you can use for your and your team’s ideation sessions. You’re also going to learn how to plan and time-manage a creative process effectively.

Most of us need to be creative in our work regardless of if we design user interfaces, write content for a website, work out appropriate workflows for an organization or program new algorithms for system backend. However, we all get those times when the creative step, which we so desperately need, simply does not come. That can seem scary—but trust us when we say that anyone can learn how to be creative­ on demand . This course will teach you ways to break the impasse of the empty page. We'll teach you methods which will help you find novel and useful solutions to a particular problem, be it in interaction design, graphics, code or something completely different. It’s not a magic creativity machine, but when you learn to put yourself in this creative mental state, new and exciting things will happen.

In the “Build Your Portfolio: Ideation Project” , you’ll find a series of practical exercises which together form a complete ideation project so you can get your hands dirty right away. If you want to complete these optional exercises, you will get hands-on experience with the methods you learn and in the process you’ll create a case study for your portfolio which you can show your future employer or freelance customers.

Your instructor is Alan Dix . He’s a creativity expert, professor and co-author of the most popular and impactful textbook in the field of Human-Computer Interaction. Alan has worked with creativity for the last 30+ years, and he’ll teach you his favorite techniques as well as show you how to make room for creativity in your everyday work and life.

You earn a verifiable and industry-trusted Course Certificate once you’ve completed the course. You can highlight it on your resume , your LinkedIn profile or your website .

All open-source articles on Creative Problem Solving

10 simple ideas to get your creative juices flowing.

• 4 years ago

Open Access—Link to us!

We believe in Open Access and the  democratization of knowledge . Unfortunately, world-class educational materials such as this page are normally hidden behind paywalls or in expensive textbooks.

If you want this to change , cite this page , link to us, or join us to help us democratize design knowledge !

Privacy Settings

Our digital services use necessary tracking technologies, including third-party cookies, for security, functionality, and to uphold user rights. Optional cookies offer enhanced features, and analytics.

Experience the full potential of our site that remembers your preferences and supports secure sign-in.

Governs the storage of data necessary for maintaining website security, user authentication, and fraud prevention mechanisms.

Enhanced Functionality

Saves your settings and preferences, like your location, for a more personalized experience.

Referral Program

We use cookies to enable our referral program, giving you and your friends discounts.

Error Reporting

We share user ID with Bugsnag and NewRelic to help us track errors and fix issues.

Optimize your experience by allowing us to monitor site usage. You’ll enjoy a smoother, more personalized journey without compromising your privacy.

Analytics Storage

Collects anonymous data on how you navigate and interact, helping us make informed improvements.

Differentiates real visitors from automated bots, ensuring accurate usage data and improving your website experience.

Lets us tailor your digital ads to match your interests, making them more relevant and useful to you.

Advertising Storage

Stores information for better-targeted advertising, enhancing your online ad experience.

Personalization Storage

Permits storing data to personalize content and ads across Google services based on user behavior, enhancing overall user experience.

Advertising Personalization

Allows for content and ad personalization across Google services based on user behavior. This consent enhances user experiences.

Enables personalizing ads based on user data and interactions, allowing for more relevant advertising experiences across Google services.

Receive more relevant advertisements by sharing your interests and behavior with our trusted advertising partners.

Enables better ad targeting and measurement on Meta platforms, making ads you see more relevant.

Allows for improved ad effectiveness and measurement through Meta’s Conversions API, ensuring privacy-compliant data sharing.

LinkedIn Insights

Tracks conversions, retargeting, and web analytics for LinkedIn ad campaigns, enhancing ad relevance and performance.

LinkedIn CAPI

Enhances LinkedIn advertising through server-side event tracking, offering more accurate measurement and personalization.

Google Ads Tag

Tracks ad performance and user engagement, helping deliver ads that are most useful to you.

Share Knowledge, Get Respect!

or copy link

Cite according to academic standards

Simply copy and paste the text below into your bibliographic reference list, onto your blog, or anywhere else. You can also just hyperlink to this page.

New to UX Design? We’re Giving You a Free ebook!

Download our free ebook The Basics of User Experience Design to learn about core concepts of UX design.

In 9 chapters, we’ll cover: conducting user interviews, design thinking, interaction design, mobile UX design, usability, UX research, and many more!

Critical thinking vs Creative thinking

Both critical thinking and creative thinking are used for solving problems , only in different ways. For critical thinking, the process is structured and methodical. For creative thinking, the process is fluid and somewhat experimental. Both thinking strategies are useful, with neither being innately superior to the other and in some unexpected ways even being linked. Now without further ado, let us explore the various components of critical thinking and creative thinking.

The Intersection of Critical Thinking & Creative Thinking

Critical thinking:.

The unbiased system attempts to remove all possible biases from thinking. Everybody has some form of bias or another. Perhaps a personal bias that one has towards someone or something. Or be it a more ethnocentric bias that prevents an individual from being able to see past the beliefs instilled in them by their culture. The unbiased analysis aims to view things from an objective instead of a subjective stand-point.

The rationality system is based on obtaining rationality, which can be defined as one being agreeable to reason. What is reason? In philosophical terms, reason is the ability to make sense of the world around us through the application of logic. Logic is a key tenet of the three systems and the cornerstone of critical thinking.

When examining deductive arguments, we begin by not looking at the truth value of the premises, but if they lead to the conclusion in a coherent manner. If they do not then the argument is deemed invalid and unsound. If the argument is deemed valid we then examine the truth value of the premises. If true, then the argument is sound, if they are not true then the argument is still valid but unsound.

Abductive arguments are drawn from the heuristic technique. The heuristic technique entails non-optimal problem-solving solutions, but are none the less sufficient for immediate decisions and approximations. Abductive reasoning includes such tactics as making an educated guess, following the general rule of thumb, or simple trial and error.

Creative thinking:

Creativity itself is the process where something truly new, but also valuable is formed. Be it a new idea, invention, or piece of art. Unlike logical thinking, there is no stringent set of rules or guidelines for how to undergo creative thinking. The process itself isn’t even entirely understood and there is much speculation and theorizing as to how creative thinking works, with no theory currently set in stone. This makes it a little more challenging to explain how to become a creative thinker. In attempting to do so we will go over some general principles of creative thinking and theories that may explain it.

Creative thinking has been hypothesized by some scientists as being a part of the evolutionary process. Some scientists think that by thinking of things in abstract terms we were better able to come up with new and innovative solutions in changing environments. Various scientists and academics have attempted to map out the process of creative thinking, one popular theory being largely developed by the psychologist J.P Guilford. Guilford helped develop the theory of divergent thinking.

Divergent thinking is the process some think is responsible for producing creativity and this is done by examining many possible solutions. Divergent thinking is more spontaneous and doesn’t occur in a linear manner. With divergent thinking a great many possible activities are explored over a short period of time, often with unexpected yet original connections being made. Common activities to help engage in divergent thinking are to create a list of questions, taking the time to think and meditate on ideas, artistic endeavors such as writing and drawing are also encouraged.

You may also like

Critical thinking and time-management (how to get the most important things done), critical thinking: where it went and how to bring it back, critical thinking skills for managers, how to overcome procrastination, download this free ebook.

What Is The Difference Between Critical Thinking And Creative Thinking

Annie Walls

In today's fast-paced world, problem-solving and decision-making have become essential skills in both personal and professional life. We face constant challenges, ranging from simple ones such as managing our daily schedule to complex ones such as strategic planning for a business. Two major thinking approaches that have gained significant attention in recent times are critical thinking and creative thinking. Although these two thinking styles are often viewed as opposites, they are both needed to solve problems effectively. In this article, we'll explore the key differences between critical thinking and creative thinking, their components, and the advantages of balancing both styles.

Understanding Critical Thinking

As individuals, we encounter a vast amount of information every day, and it is essential to be able to analyze and evaluate that information to make informed decisions. Critical thinking is a valuable skill that enables individuals to think logically and systematically, while also questioning assumptions and biases.

Definition of Critical Thinking

At its core, critical thinking involves analyzing information and ideas by breaking them down into smaller components to evaluate their accuracy, validity, relevance, and completeness. This process allows individuals to question assumptions, evidence, and arguments made by others and use logical reasoning to make sound decisions.

Furthermore, critical thinking involves identifying biases, stereotypes, and fallacies and preventing them from influencing our judgment. It is a systematic process that entails evaluating multiple sources of information and perspectives, weighing up evidence, and forming an informed opinion based on available facts.

Components of Critical Thinking

Critical thinking comprises several components that play a crucial role in the thinking process. According to experts, the key components of critical thinking include:

• Interpretation - understanding the meaning and significance of information. This component involves analyzing and interpreting data, identifying patterns and trends, and drawing conclusions based on the available evidence.
• Analysis - breaking down complex information into smaller parts to examine their relationships and distinctions. This component involves identifying the underlying assumptions and arguments, evaluating the evidence presented, and identifying any potential biases or fallacies.
• Evaluation - assessing the strength and validity of arguments and evidence presented. This component involves evaluating the credibility of sources, weighing up the evidence presented, and identifying any gaps in the argument.
• Inference - making logical conclusions based on available information and evidence. This component involves drawing conclusions based on the available evidence and identifying any potential implications or consequences.
• Explanation - presenting a clear and concise rationale for the conclusions drawn. This component involves communicating the results of the analysis and evaluation in a clear and concise manner.
• Self-regulation - monitoring one's own thinking process to overcome biases and errors. This component involves being aware of one's own biases and assumptions and actively seeking out new information to challenge those assumptions.

Examples of Critical Thinking in Action

Critical thinking is widely used in various domains of life, including education, healthcare, law enforcement, and business. Here are a few examples of how critical thinking is applied in practice:

• A doctor making a diagnosis based on a set of symptoms observed - A doctor uses critical thinking to analyze the symptoms presented by a patient, evaluate potential causes, and make a diagnosis based on the available evidence.
• A lawyer evaluating evidence presented in a court hearing to support their case - A lawyer uses critical thinking to evaluate the credibility of witnesses, weigh up the evidence presented, and construct a compelling argument to support their case.
• A teacher designing a lesson plan that engages students in critical thinking skills - A teacher uses critical thinking to design a lesson plan that encourages students to analyze and evaluate information, draw conclusions based on the available evidence, and communicate their findings effectively.
• A business executive analyzing market trends and customer preferences to make informed decisions - A business executive uses critical thinking to analyze market trends, evaluate customer preferences, and make informed decisions based on the available evidence.

In conclusion, critical thinking is a vital skill that enables individuals to analyze and evaluate information, draw logical conclusions, and make informed decisions. By developing critical thinking skills, individuals can overcome biases and assumptions, evaluate evidence objectively, and communicate their findings effectively.

Understanding Creative Thinking

Creativity is a fascinating and complex phenomenon that has captivated the attention of scholars, artists, and entrepreneurs for centuries. It is a multifaceted construct that involves a wide range of cognitive, affective, and behavioral processes. At its core, creativity is about generating novel and valuable ideas that have the potential to transform the world.

Definition of Creative Thinking

Creative thinking is the process of generating new ideas, possibilities, or solutions that are original, useful, and novel. It involves breaking away from traditional or conventional patterns of thinking and exploring alternative perspectives. Creative thinking is characterized by fluidity, flexibility, and originality. It is a free-flowing mindset that allows individuals to connect dissimilar ideas and develop innovative concepts that solve problems or meet needs. Creative thinking is central to innovation, invention, and entrepreneurship.

When we engage in creative thinking, we are tapping into our imagination and exploring the unknown. We are willing to take risks, challenge assumptions, and embrace ambiguity. Creative thinking is not just about coming up with wild and crazy ideas; it is about generating ideas that are both feasible and valuable.

Components of Creative Thinking

Creative thinking comprises several interrelated components, including:

• Fluency: This refers to the ability to generate a large number of ideas with ease. The more ideas we generate, the more likely we are to come up with something truly innovative.
• Flexibility: This involves considering different possibilities and perspectives. We need to be open-minded and willing to explore diverse options in order to generate truly creative ideas.
• Originality: This refers to the ability to produce unconventional and unique ideas. We need to break away from conventional thinking and explore new and uncharted territories.
• Elaboration: This involves refining and developing ideas with details and depth. We need to flesh out our ideas and explore their potential in order to turn them into reality.
• Imagery: This refers to the ability to visualize and manipulate images or metaphors to generate ideas. We can use our imagination to create mental images that inspire us and spark our creativity.
• Association: This involves connecting seemingly unrelated ideas to form new concepts. We need to be able to see the connections between different ideas and concepts in order to generate truly innovative ideas.

Examples of Creative Thinking in Action

Creative thinking is applied in various fields, such as art, design, science, and technology. Here are some examples of how creative thinking is used:

• An artist creating a new genre of art: By blending traditional and modern techniques, an artist can create a new style of art that is both unique and captivating.
• A software developer designing a user-friendly interface: By combining cutting-edge technology with user-centered design principles, a software developer can create an interface that is both intuitive and efficient.
• A chef creating a unique dish: By combining diverse ingredients and experimenting with new flavors and textures, a chef can create a dish that is both delicious and memorable.
• A scientist inventing a new process: By exploring new methods and technologies, a scientist can invent a new process that has the potential to revolutionize an industry or even change the world.

Overall, creative thinking is a powerful tool that can help us solve problems, innovate, and make a positive impact on the world. By embracing our creativity and exploring new possibilities, we can unlock our full potential and achieve great things.

Key Differences Between Critical and Creative Thinking

Critical thinking and creative thinking are two distinct modes of thinking that have different purposes, processes, and outcomes. While critical thinking is focused on evaluating and analyzing information, creative thinking is aimed at generating new and innovative ideas and solutions to problems. Let's explore the differences in more detail.

Purpose and Goals

Critical thinking is primarily aimed at evaluating, analyzing, and critiquing information critically. It involves questioning assumptions, evaluating evidence, and identifying biases and fallacies in arguments. The goal of critical thinking is to arrive at well-reasoned and informed judgments or decisions based on the available evidence.

On the other hand, creative thinking is focused on generating new and innovative ideas and solutions to problems. It involves exploring possibilities, making connections, and thinking outside the box. The goal of creative thinking is to come up with novel and useful ideas that can lead to innovation and change.

Process and Approach

Critical thinking requires a systematic and rigorous process of analysis and evaluation based on available evidence and data. It involves breaking down complex information into its component parts, examining each part critically, and evaluating the evidence and arguments presented. Critical thinking requires a structured approach that involves identifying and analyzing arguments, evaluating evidence, and drawing conclusions based on the available information.

Creative thinking, on the other hand, involves a free-flowing process of divergent thinking that encourages unconventional ideas and connections. It involves exploring different perspectives, generating multiple ideas, and making unexpected connections between seemingly unrelated concepts. Creative thinking requires a more open and exploratory approach that involves brainstorming, visualizing, and associating ideas.

Skills and Abilities Involved

Critical thinking is associated with skills such as analysis, inference, evaluation, and reasoning. It requires the ability to identify and evaluate evidence, recognize biases and assumptions, and draw well-reasoned conclusions based on the available information. Critical thinking also involves the ability to communicate effectively, both orally and in writing.

Creative thinking requires skills such as imagination, brainstorming, visualization, and association. It involves the ability to generate new and original ideas, make connections between seemingly unrelated concepts, and explore different perspectives. Creative thinking also requires the ability to communicate ideas effectively, both orally and in writing.

In conclusion, critical thinking and creative thinking are two distinct modes of thinking that have different purposes, processes, and outcomes. While critical thinking is focused on evaluating and analyzing information, creative thinking is aimed at generating new and innovative ideas and solutions to problems. Both modes of thinking are important for success in today's complex and rapidly changing world.

The Importance of Balancing Critical and Creative Thinking

Benefits of combining both types of thinking.

The integration of critical and creative thinking leads to better problem solving, decision making, and innovation by combining analysis and creativity. The following are some benefits of balancing these thinking styles:

• Increase in productivity and efficiency
• Improved communication and teamwork skills
• Better problem-solving and decision-making abilities
• Enhancement of individual and organizational creativity and innovation

Strategies for Developing a Balanced Thinking Approach

Developing a balanced thinking approach requires a conscious effort to integrate critical and creative thinking strategies. Some practical ways of achieving this are:

• Practicing active listening and asking thoughtful questions to clarify and evaluate information
• Encouraging brainstorming sessions that involve diverse perspectives and ideas
• Challenging personal assumptions and biases and adopting a growth mindset
• Using visualization techniques to generate creative solutions to problems

Real-World Applications of Balanced Thinking

The application of a balanced thinking approach leads to better decision-making processes and outcomes. Here are some real-world examples:

• A company using a combination of critical analysis and creativity to develop new products and marketing strategies.
• An individual using critical thinking skills to analyze career opportunities and creative thinking skills to identify alternative paths.
• A teacher using critical analysis to evaluate student's work and creative thinking to design engaging lessons that foster innovation and growth.

To sum up, critical thinking and creative thinking are two essential thinking skills needed for successful problem solving and decision making. While critical thinking involves the systematic evaluation of information and arguments, creative thinking is focused on generating novel and innovative ideas and solutions. Balancing these thinking styles results in enhanced productivity, better communication, and more creative and effective problem-solving. By combining critical and creative thinking strategies, individuals can achieve a balanced thinking approach that leads to better decision-making processes and outcomes.

Popular Posts

Robert hannigan – the power of neurodiversity in innovation, cybersecurity, gchq and counter-intelligence #342.

Explore key insights on intelligence and decision-making from Professor Sir David Omand’s book, focusing on critical thinking and creativity.

Sam Dixon of Womble Bond Dickinson, The Evolving Role of Lawyers in the AI Era #341

John craske of cms, collaboration between humans and machines in the legal industry #340, jd meier of microsoft, productivity strategies for success #339, sir david omand, author of how spies think – 10 lessons in critical thinking #338, meilleur conférencier principal en teambuilding.

Les conférences virtuelles et les sommets peuvent être des moyens très efficaces pour inspirer, informer

James is a top motivational keynote speaker who is booked as a creativity and innovation keynote speaker, AI speaker , sustainability speaker and leadership speaker . Recent destinations include: Dubai , Abu Dhabi , Orlando , Las Vegas , keynote speaker London , Barcelona , Bangkok , Miami , Berlin , Riyadh , New York , Zurich , motivational speaker Paris , Singapore and San Francisco

Latest News

• 415.800.3059
• [email protected]
• Media Interviews
• Meeting Planners
• Terms of Use
• Privacy Policy
• Cookie Policy

FIND ME ON SOCIAL

© 2024 James Taylor DBA P3 Music Ltd.

• Business Essentials
• Leadership & Management
• Credential of Leadership, Impact, and Management in Business (CLIMB)
• Entrepreneurship & Innovation
• Digital Transformation
• Finance & Accounting
• Business in Society
• For Organizations
• Support Portal
• Media Coverage
• Founding Donors
• Leadership Team

• Harvard Business School →
• HBS Online →
• Business Insights →

Business Insights

Harvard Business School Online's Business Insights Blog provides the career insights you need to achieve your goals and gain confidence in your business skills.

• Career Development
• Communication
• Decision-Making
• Earning Your MBA
• Negotiation
• News & Events
• Productivity
• Staff Spotlight
• Student Profiles
• Work-Life Balance
• AI Essentials for Business
• Alternative Investments
• Business Analytics
• Business Strategy
• Business and Climate Change
• Creating Brand Value
• Design Thinking and Innovation
• Digital Marketing Strategy
• Disruptive Strategy
• Economics for Managers
• Entrepreneurship Essentials
• Financial Accounting
• Global Business
• Launching Tech Ventures
• Leadership Principles
• Leadership, Ethics, and Corporate Accountability
• Leading Change and Organizational Renewal
• Leading with Finance
• Management Essentials
• Negotiation Mastery
• Organizational Leadership
• Power and Influence for Positive Impact
• Strategy Execution
• Sustainable Business Strategy
• Sustainable Investing
• Winning with Digital Platforms

What Is Design Thinking & Why Is It Important?

• 18 Jan 2022

In an age when innovation is key to business success and growth, you’ve likely come across the term “design thinking.” Perhaps you’ve heard it mentioned by a senior leader as something that needs to be utilized more, or maybe you’ve seen it on a prospective employee's resume.

While design thinking is an ideology based on designers’ workflows for mapping out stages of design, its purpose is to provide all professionals with a standardized innovation process to develop creative solutions to problems—design-related or not.

Why is design thinking needed? Innovation is defined as a product, process, service, or business model featuring two critical characteristics: novel and useful. Yet, there’s no use in creating something new and novel if people won’t use it. Design thinking offers innovation the upgrade it needs to inspire meaningful and impactful solutions.

But what is design thinking, and how does it benefit working professionals?

What Is Design Thinking?

Design thinking is a mindset and approach to problem-solving and innovation anchored around human-centered design . While it can be traced back centuries—and perhaps even longer—it gained traction in the modern business world after Tim Brown, CEO and president of design company IDEO, published an article about it in the Harvard Business Review .

Design thinking is different from other innovation and ideation processes in that it’s solution-based and user-centric rather than problem-based. This means it focuses on the solution to a problem instead of the problem itself.

For example, if a team is struggling with transitioning to remote work, the design thinking methodology encourages them to consider how to increase employee engagement rather than focus on the problem (decreasing productivity).

The essence of design thinking is human-centric and user-specific. It’s about the person behind the problem and solution, and requires asking questions such as “Who will be using this product?” and “How will this solution impact the user?”

The first, and arguably most important, step of design thinking is building empathy with users. By understanding the person affected by a problem, you can find a more impactful solution. On top of empathy, design thinking is centered on observing product interaction, drawing conclusions based on research, and ensuring the user remains the focus of the final implementation.

The Four Phases of Innovation

So, what does design thinking entail? There are many models of design thinking that range from three to seven steps.

In the online course Design Thinking and Innovation , Harvard Business School Dean Srikant Datar leverages a four-phase innovation framework. The phases venture from concrete to abstract thinking and back again as the process loops, reverses, and repeats. This is an important balance because abstract thinking increases the likelihood that an idea will be novel. It’s essential, however, to anchor abstract ideas in concrete thinking to ensure the solution is valid and useful.

Here are the four phases for effective innovation and, by extension, design thinking.

The first phase is about narrowing down the focus of the design thinking process. It involves identifying the problem statement to come up with the best outcome. This is done through observation and taking the time to determine the problem and the roadblocks that prevented a solution in the past.

Various tools and frameworks are available—and often needed—to make concrete observations about users and facts gathered through research. Regardless of which tools are implemented, the key is to observe without assumptions or biased expectations.

Once findings from your observations are collected, the next step is to shape insights by framing those observations. This is where you can venture into the abstract by reframing the problem in the form of a statement or question.

Once the problem statement or question has been solidified—not finalized—the next step is ideation. You can use a tool such as systematic inventive thinking (SIT) in this stage, which is useful for creating an innovative process that can be replicated in the future.

The goal is to ultimately overcome cognitive fixedness and devise new and innovative ideas that solve the problems you identified. Continue to actively avoid assumptions and keep the user at the forefront of your mind during ideation sessions.

The third phase involves developing concepts by critiquing a range of possible solutions. This includes multiple rounds of prototyping, testing, and experimenting to answer critical questions about a concept’s viability.

Remember: This step isn’t about perfection, but rather, experimenting with different ideas and seeing which parts work and which don’t.

4. Implement

The fourth and final phase, implementation, is when the entire process comes together. As an extension of the develop phase, implementation starts with testing, reflecting on results, reiterating, and testing again. This may require going back to a prior phase to iterate and refine until you find a successful solution. Such an approach is recommended because design thinking is often a nonlinear, iterative process.

In this phase, don’t forget to share results with stakeholders and reflect on the innovation management strategies implemented during the design thinking process. Learning from experience is an innovation process and design thinking project all its own.

Check out the video about the design thinking process below, and subscribe to our YouTube channel for more explainer content!

Why Design Thinking Skills Matter

The main value of design thinking is that it offers a defined process for innovation. While trial and error is a good way to test and experiment what works and what doesn’t, it’s often time-consuming, expensive, and ultimately ineffective. On the other hand, following the concrete steps of design thinking is an efficient way to develop new, innovative solutions.

On top of a clear, defined process that enables strategic innovation, design thinking can have immensely positive outcomes for your career—in terms of both advancement and salary.

As of December 2021, the most common occupations requiring design thinking skills were:

• Marketing managers
• Industrial engineers
• Graphic designers
• Software developers
• General and operations managers
• Management analysts
• Personal service managers
• Architectural and engineering managers
• Computer and information systems managers

In addition, jobs that require design thinking statistically have higher salaries. Take a marketing manager position, for example. The median annual salary is $107,900. Marketing manager job postings that require design thinking skills, however, have a median annual salary of $133,900—a 24 percent increase.

Overall, businesses are looking for talent with design thinking skills. As of November 2021, there were 29,648 job postings in the United States advertising design thinking as a necessary skill—a 153 percent increase from November 2020, and a 637 percent increase from November 2017.

As businesses continue to recognize the need for design thinking and innovation, they’ll likely create more demand for employees with those skills.

Learning Design Thinking

Design thinking is an extension of innovation that allows you to design solutions for end users with a single problem statement in mind. It not only imparts valuable skills but can help advance your career.

It’s also a collaborative endeavor that can only be mastered through practice with peers. As Datar says in the introduction to Design Thinking and Innovation : “Just as with learning how to swim, the best way to practice is to jump in and try.”

If you want to learn design thinking, take an active role in your education. Start polls, problem-solving exercises, and debates with peers to get a taste of the process. It’s also important to seek out diverse viewpoints to prepare yourself for the business world.

In addition, if you’re considering adding design thinking to your skill set, think about your goals and why you want to learn about it. What else might you need to be successful?

You might consider developing your communication, innovation, leadership, research, and management skills, as those are often listed alongside design thinking in job postings and professional profiles.

You may also notice skills like agile methodology, user experience, and prototyping in job postings, along with non-design skills, such as product management, strategic planning, and new product development.

Is Design Thinking Right for You?

There are many ways to approach problem-solving and innovation. Design thinking is just one of them. While it’s beneficial to learn how others have approached problems and evaluate if you have the same tools at your disposal, it can be more important to chart your own course to deliver what users and customers truly need.

You can also pursue an online course or workshop that dives deeper into design thinking methodology. This can be a practical path if you want to improve your design thinking skills or require a more collaborative environment.

Are you ready to develop your design thinking skills? Explore our online course Design Thinking and Innovation to discover how to leverage fundamental design thinking principles and innovative problem-solving tools to address business challenges.

About the Author

Want to create or adapt books like this? Learn more about how Pressbooks supports open publishing practices.

Part Two: You are the President and CEO of You

Thinking Critically and Creatively

Dr. andrew robert baker.

Critical and creative thinking skills are perhaps the most fundamental skills involved in making judgments and solving problems. They are some of the most important skills I have ever developed. I use them everyday and continue to work to improve them both.

The ability to think critically about a matter—to analyze a question, situation, or problem down to its most basic parts—is what helps us evaluate the accuracy and truthfulness of statements, claims, and information we read and hear. It is the sharp knife that, when honed, separates fact from fiction, honesty from lies, and the accurate from the misleading. We all use this skill to one degree or another almost every day. For example, we use critical thinking every day as we consider the latest consumer products and why one particular product is the best among its peers. Is it a quality product because a celebrity endorses it? Because a lot of other people may have used it? Because it is made by one company versus another? Or perhaps because it is made in one country or another? These are questions representative of critical thinking.

The academic setting demands more of us in terms of critical thinking than everyday life. It demands that we evaluate information and analyze a myriad of issues. It is the environment where our critical thinking skills can be the difference between success and failure. In this environment we must consider information in an analytical, critical manner. We must ask questions—What is the source of this information? Is this source an expert one and what makes it so? Are there multiple perspectives to consider on an issue? Do multiple sources agree or disagree on an issue? Does quality research substantiate information or opinion? Do I have any personal biases that may affect my consideration of this information? It is only through purposeful, frequent, intentional questioning such as this that we can sharpen our critical thinking skills and improve as students, learners, and researchers. Developing my critical thinking skills over a twenty year period as a student in higher education enabled me to complete a quantitative dissertation, including analyzing research and completing statistical analysis, and earning my Ph.D. in 2014.

While critical thinking analyzes information and roots out the true nature and facets of problems, it is creative thinking that drives progress forward when it comes to solving these problems. Exceptional creative thinkers are people that invent new solutions to existing problems that do not rely on past or current solutions. They are the ones who invent solution C when everyone else is still arguing between A and B. Creative thinking skills involve using strategies to clear the mind so that our thoughts and ideas can transcend the current limitations of a problem and allow us to see beyond barriers that prevent new solutions from being found.

Brainstorming is the simplest example of intentional creative thinking that most people have tried at least once. With the quick generation of many ideas at once we can block-out our brain’s natural tendency to limit our solution-generating abilities so we can access and combine many possible solutions/thoughts and invent new ones. It is sort of like sprinting through a race’s finish line only to find there is new track on the other side and we can keep going, if we choose. As with critical thinking, higher education both demands creative thinking from us and is the perfect place to practice and develop the skill. Everything from word problems in a math class, to opinion or persuasive speeches and papers, call upon our creative thinking skills to generate new solutions and perspectives in response to our professor’s demands. Creative thinking skills ask questions such as—What if? Why not? What else is out there? Can I combine perspectives/solutions? What is something no one else has brought-up? What is being forgotten/ignored? What about ______? It is the opening of doors and options that follows problem-identification.

Consider an assignment that required you to compare two different authors on the topic of education and select and defend one as better. Now add to this scenario that your professor clearly prefers one author over the other. While critical thinking can get you as far as identifying the similarities and differences between these authors and evaluating their merits, it is creative thinking that you must use if you wish to challenge your professor’s opinion and invent new perspectives on the authors that have not previously been considered.

So, what can we do to develop our critical and creative thinking skills? Although many students may dislike it, group work is an excellent way to develop our thinking skills. Many times I have heard from students their disdain for working in groups based on scheduling, varied levels of commitment to the group or project, and personality conflicts too, of course. True—it’s not always easy, but that is why it is so effective. When we work collaboratively on a project or problem we bring many brains to bear on a subject. These different brains will naturally develop varied ways of solving or explaining problems and examining information. To the observant individual we see that this places us in a constant state of back and forth critical/creative thinking modes.

For example, in group work we are simultaneously analyzing information and generating solutions on our own, while challenging other’s analyses/ideas and responding to challenges to our own analyses/ideas. This is part of why students tend to avoid group work—it challenges us as thinkers and forces us to analyze others while defending ourselves, which is not something we are used to or comfortable with as most of our educational experiences involve solo work. Your professors know this—that’s why we assign it—to help you grow as students, learners, and thinkers!

Foundations of Academic Success: Words of Wisdom Copyright © 2015 by Thomas Priester is licensed under a Creative Commons Attribution 4.0 International License , except where otherwise noted.

What Is the Difference Between Analytical and Creative Problem Solving?

Anne pyburn craig.

Analytical and creative problem solving abilities rely on different skill sets. Analytical thinking is also referred to as logical thinking, while creative thinking can also be called "lateral" thinking. Sometimes the difference is described in terms of left-brain, or analytical, and right brain, or creative, thinking. Analytical thinking is useful in solving convergent problems, ones to which there is only one correct answer. Creative thinking works better in solving divergent problems, where there may be many or even infinite solutions. Many problems respond best to a mix of both.

Explore this article

• Problems That Require Analytical Thinking
• Analytical Problem Solving Skills
• Problems That Require Creative Thinking
• Creative Thinking Skills

1 Problems That Require Analytical Thinking

Analytical thinking is essential in solving logistical problems, such as, "What's the best way to ship this load of widgets to Muncie, Indiana?" or "How do I make a week's worth of meals on my limited food budget?" For either of these, you will need a knowledge of facts and the ability to put them together in the right way: shipping prices and speeds, for example, or what ingredients you'll need to make a meatloaf.

2 Analytical Problem Solving Skills

Analytical thinking depends on the ability to recall or research facts and figures that are relevant to the problem. Being able to categorize these facts, use the right ones in the right ways and think logically about the probable outcomes of various options is key to analytical problem solving. So is the ability to narrow down options, eliminating irrelevant data or unworkable choices. Analytical problem solving also calls for the ability to observe trends and tendencies and use what has happened in the past to predict probable outcomes.

3 Problems That Require Creative Thinking

Creative thinking is essential in finding solutions to problems with complex or abstract elements, such as, "What's the best way to govern a city?" or "Why don't my children get along better?" These kinds of problems have no correct answer that applies in every situation, so generating creative ideas is a must. Most of the bigger questions in life call for at least some creative thinking.

4 Creative Thinking Skills

Creative thinking involves the ability to see not just what is evident but also what may be the missing pieces of a solution. One common creative problem-solving technique is brainstorming, in which numerous ideas are collected before any of them are criticized or discarded. Another important skill in creative problem solving is the ability to suspend judgment and follow a train of thought imaginatively to see where it may lead. And sometimes it is helpful to lay a problem aside, allowing its elements to percolate in the subconscious mind, and return to it later to see what new angles or solutions may emerge.

• 1 ITS Education: The Skills of Problem Solving
• 2 Ready To Manage: How Is Critical Thinking Different From Analytical or Lateral Thinking?

About the Author

Anne Pyburn Craig has written for a range of regional and local publications ranging from in-depth local investigative journalism to parenting, business, real estate and green building publications. She frequently writes tourism and lifestyle articles for chamber of commerce publications and is a respected book reviewer.

Related Articles

Critical Thinking vs. Non-Critical Thinking

Techniques Used to Solve Ethical Dilemmas

What Are the Four Forms of Critical Thinking and Writing?

Creative and Critical Thinking Skills

Remanufactured Vs. Reconditioned

What Are Linear & Nonlinear Thinkers?

How to Make Chrome Refresh Automatically

Enzyme Activity in Photosynthesis

What Is the Chemical Formula for Magnesium Nitrate...

Quick Math Facts for the 4th Grade

Good Ways to Start a Comparison and Contrast Essay

What Is the Chemical Formula for Bromine Reacting With...

How to Write a Conclusion

Differences Between Analytical & Critical Thinking

How to Identify an Ethical Dilemma

Problem-Solving Skills for Adults

How to Solve the Ethical Dilemma of Stealing a Loaf...

Signs of Intelligence

How to Identify the 6 Types of Chemical Reactions

What Are the Characteristics of Critical & Creative...

Regardless of how old we are, we never stop learning. Classroom is the educational resource for people of all ages. Whether you’re studying times tables or applying to college, Classroom has the answers.

• Accessibility
• Terms of Use
• Privacy Policy
• Copyright Policy
• Manage Preferences

© 2020 Leaf Group Ltd. / Leaf Group Media, All Rights Reserved. Based on the Word Net lexical database for the English Language. See disclaimer .

Creative Problem Solving in Large Language and Vision Models – What Would it Take?

We advocate for a strong integration of Computational Creativity (CC) with research in large language and vision models (LLVMs) to address a key limitation of these models, i.e., creative problem solving. We present preliminary experiments showing how CC principles can be applied to address this limitation. Our goal is to foster discussions on creative problem solving in LLVMs and CC at prestigious ML venues.

Lakshmi Nair Georgia Institute of Technology Atlanta, GA, USA                        Evana Gizzi Tufts University Medford, MA, USA                        Jivko Sinapov Tufts University Medford, MA, USA

1 Introduction

Creativity is “ …the ability to come up with an idea which, relative to the pre-existing domain-space in one’s mind, one could not have had before. Whether any other person (or system) has already come up with it on an earlier occasion is irrelevant. ” Boden ( 1998 ) , p.216. For artificial agents, Computational Creativity (CC) is a multi-disciplinary field (spanning Philosophy, Psychology, Neuroscience, and Computer Science) that seeks to develop computational methods capable of generating creative outcomes reminiscent of creative processes in humans Gizzi et al. ( 2022 ) . Within CC, creative problem solving is a sub-area that requires an agent to discover – from its perspective – novel and previously unseen ways to accomplish a task. For example, in the absence of a ladle to scoop ingredients, an agent might creatively choose to substitute a bowl in place of the ladle. In this sense, creative problem solving encompasses creativity that is specifically task-oriented , as opposed to the generation of creative artifacts e.g., music or images.

While recent state-of-the-art large language models (LLMs) and vision-language models (VLMs) have demonstrated competency in artistic endeavours Rombach et al. ( 2021 ); Copet et al. ( 2023 ) , creative problem solving continues to be a shortcoming of these models (we use LLVM to denote the umbrella of both LLMs and VLMs). For instance, in Bubeck et al. ( 2023 ) , the authors point out that “discontinuous tasks” that require a certain “Eureka” idea, i.e., creative problem solving, is currently a limitation of models like GPT-4. Similar observations have been made in follow up work showing that state-of-the-art LLMs inherently possess poor creative problem solving capabilities compared to humans Tian et al. ( 2023 ); Naeini et al. ( 2023 ) . Given this obvious limitation, ongoing research in Machine Learning should seek to address the gap between LLVMs and creative problem solving, to further enhance the intelligent capabilities of these models. As defined in prior work, “ Intelligence is the ability to work and adapt to the environment with insufficient knowledge and resources. ” Pennachin and Goertzel ( 2007 ) , p.10. Demonstrated in hallmark examples of human ingenuity, like the makeshift C ⁢ O 2 𝐶 subscript 𝑂 2 CO_{2} italic_C italic_O start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT filter built onboard the Apollo-13 Cass ( 2005 ) , or the makeshift medical devices used to offset equipment shortages during COVID-19 Turner et al. ( 2020 ) , creative problem solving is especially important when dealing with resource-critical scenarios. Since humans may tend to “choke” under high pressure situations DeCaro et al. ( 2011 ) often limiting their CPS skills, autonomous agents equipped with LLVMs that have similar capabilities would be highly assistive and transformative to humans in high-stake environments. These include situations like rescue missions BBC ( 2012 ) or autonomous operation in human-inaccessible environments (e.g., space or underwater exploration) with limited resources Atkeson et al. ( 2018 ) . However, the exceptional degree of creative problem solving necessary for such assistance remains beyond the scope of LLVMs today, limiting their intelligence (See Appx. B.1 ).

We believe that a discussion of Computational Creativity is essential to addressing this limitation. It is our position that Machine Learning and Computational Creativity should be strongly integrated in research to enable effective creative problem solving in LLVMs and push the frontiers of their ingenuity.

2 Two Cultures Problem: Why does CC not receive a wider reception in ML?

Even though creative problem solving (CPS) is a shortcoming of existing LLVMs, Computational Creativity seldom finds its way into mainstream ML research. We believe this discrepancy aligns with the “two cultures” problem Hammond et al. ( 2013 ) (also corroborated in Van Heerden and Bas ( 2021 ); Lahikainen et al. ( 2024 ) ), and is motivated by three aspects of CC literature as it relates to creative problem solving: a) the lack of a precise definition of CPS makes it challenging to identify how existing approaches in LLVMs are deficient in CPS skills; b) the somewhat “abstract” computational descriptions of CPS in Computational Creativity is challenging to connect to practical algorithms in LLVMs; and c) the lack of standardized benchmarks make it harder to evaluate LLVMs for CPS. In our discussions relating to a) in Section 3.1 , b) in Section 4 , and c) Section 5 , we hope to address these gaps and encourage the ML community to think about how LLVMs can be augmented with creative problem solving skills through a deeper discussion of Computational Creativity.

To emphasize the applicability of principles from CC for creative problem solving in LLVMs, we discuss the seminal work of Margaret A. Boden from CC literature that introduces three forms of creativity, namely, “ exploratory ”, “ combinational ”, and “ transformational ” Boden ( 1998 ) . Prior work has discussed the extension of Boden’s forms of creativity to creative problem solving in AI Gizzi et al. ( 2022 ) , however, their work does not include recent advances in LLVMs nor how Boden’s principles can be extended to specific approaches for LLVMs.

Ongoing discussions by leading ML experts like Dr. Shane Legg, co-founder of DeepMind, have suggested that “search” could help such models perform creative problem solving, quote, “ … these foundational models are world models of a kind, and to do really creative problem solving, you need to start searching ” Patel ( 2023 ) . There has also been speculation that OpenAI’s Q ∗ superscript 𝑄 Q^{*} italic_Q start_POSTSUPERSCRIPT ∗ end_POSTSUPERSCRIPT search (described as a “significant breakthrough” in popular media) could be targeting a similar approach Wang ( 2023 ); Anna Tong and Hu ( 2023 ) . Interestingly, we note that “search” as described here, can be linked to Boden’s proposed “exploratory” approach (Section 4.1.1 ). However, in Section 4 , we posit that “combinational” and “transformational” modes should be equally emphasized to achieve creative problem solving in LLVMs.

Although we choose to expand on Boden’s work as the focal point to drive our arguments in the main paper, it is not the only theory in CC that is relevant to this discussion. For completeness, we elaborate on additional CC theories and their applicability to creative problem solving in LLVMs in Appx. B .

3 From Task Planning to Creative Problem Solving

Creative problem solving can be broadly described as the process through which agents discover novel ways of accomplishing a task that, prior to the discovery, was unsolvable. Computationally, creative problem solving can be achieved through planning, learning, or hybrid approaches Gizzi et al. ( 2022 ) . Following a review of the different definitions of creative problem solving that have been proposed (Appx. A ), we believe the following most closely connects to existing formalisms in ML.

3.1 Definition of Creative Problem Solving

Gizzi et al. ( 2022 ) define the notion of a concept , as a state (of the environment and/or agent) or action. More generally, the authors denote C X subscript 𝐶 𝑋 C_{X} italic_C start_POSTSUBSCRIPT italic_X end_POSTSUBSCRIPT as the set of all concepts relating to X 𝑋 X italic_X ( X 𝑋 X italic_X denotes environment states S 𝑆 S italic_S or actions A 𝐴 A italic_A ). Hence, C S subscript 𝐶 𝑆 C_{S} italic_C start_POSTSUBSCRIPT italic_S end_POSTSUBSCRIPT denotes the set of all environmental states, and C A subscript 𝐶 𝐴 C_{A} italic_C start_POSTSUBSCRIPT italic_A end_POSTSUBSCRIPT denotes the set of agent actions. Formally, the authors state their definition as (Page 7, (Gizzi et al., 2022 ) ):

Given an un-achievable goal due to an insufficient conceptual space, CPS refers to the process by which the agent discovers a new conceptual space C X ′ ⊈ C X not-subset-of-nor-equals subscript superscript 𝐶 ′ 𝑋 subscript 𝐶 𝑋 C^{\prime}_{X}\nsubseteq C_{X} italic_C start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_X end_POSTSUBSCRIPT ⊈ italic_C start_POSTSUBSCRIPT italic_X end_POSTSUBSCRIPT , such that C X ′ = f ⁢ ( C X ) subscript superscript 𝐶 ′ 𝑋 𝑓 subscript 𝐶 𝑋 C^{\prime}_{X}=f(C_{X}) italic_C start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_X end_POSTSUBSCRIPT = italic_f ( italic_C start_POSTSUBSCRIPT italic_X end_POSTSUBSCRIPT ) is the result of applying some function f 𝑓 f italic_f on the current conceptual space, enabling the agent to solve the previously unsolvable task by using C X ′ subscript superscript 𝐶 ′ 𝑋 C^{\prime}_{X} italic_C start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_X end_POSTSUBSCRIPT .

As a simplified example, let us assume a robot that has a goal G 𝐺 G italic_G of transferring beans from a jar to a cooker: G = 𝐺 absent G= italic_G = { i ⁢ n 𝑖 𝑛 in italic_i italic_n (beans, cooker)}. Here, the initial state is defined as C S = subscript 𝐶 𝑆 absent C_{S}= italic_C start_POSTSUBSCRIPT italic_S end_POSTSUBSCRIPT = { i ⁢ n 𝑖 𝑛 in italic_i italic_n (beans, jar), h ⁢ a ⁢ s ⁢ C ⁢ o ⁢ n ⁢ t ⁢ a ⁢ i ⁢ n ⁢ a ⁢ b ⁢ i ⁢ l ⁢ i ⁢ t ⁢ y ℎ 𝑎 𝑠 𝐶 𝑜 𝑛 𝑡 𝑎 𝑖 𝑛 𝑎 𝑏 𝑖 𝑙 𝑖 𝑡 𝑦 hasContainability italic_h italic_a italic_s italic_C italic_o italic_n italic_t italic_a italic_i italic_n italic_a italic_b italic_i italic_l italic_i italic_t italic_y (spoon)}. Let the actions be defined as C A = subscript 𝐶 𝐴 absent C_{A}= italic_C start_POSTSUBSCRIPT italic_A end_POSTSUBSCRIPT = { s ⁢ c ⁢ o ⁢ o ⁢ p 𝑠 𝑐 𝑜 𝑜 𝑝 scoop italic_s italic_c italic_o italic_o italic_p (beans, X 𝑋 X italic_X , l ⁢ o ⁢ c s 𝑙 𝑜 subscript 𝑐 𝑠 loc_{s} italic_l italic_o italic_c start_POSTSUBSCRIPT italic_s end_POSTSUBSCRIPT , l ⁢ o ⁢ c d 𝑙 𝑜 subscript 𝑐 𝑑 loc_{d} italic_l italic_o italic_c start_POSTSUBSCRIPT italic_d end_POSTSUBSCRIPT )}, where, X 𝑋 X italic_X refers to an object that satisfies h ⁢ a ⁢ s ⁢ C ⁢ o ⁢ n ⁢ t ⁢ a ⁢ i ⁢ n ⁢ a ⁢ b ⁢ i ⁢ l ⁢ i ⁢ t ⁢ y ⁢ ( ⋅ ) ℎ 𝑎 𝑠 𝐶 𝑜 𝑛 𝑡 𝑎 𝑖 𝑛 𝑎 𝑏 𝑖 𝑙 𝑖 𝑡 𝑦 ⋅ hasContainability(\cdot) italic_h italic_a italic_s italic_C italic_o italic_n italic_t italic_a italic_i italic_n italic_a italic_b italic_i italic_l italic_i italic_t italic_y ( ⋅ ) (e.g., spoon), to scoop beans from l ⁢ o ⁢ c s 𝑙 𝑜 subscript 𝑐 𝑠 loc_{s} italic_l italic_o italic_c start_POSTSUBSCRIPT italic_s end_POSTSUBSCRIPT to l ⁢ o ⁢ c d 𝑙 𝑜 subscript 𝑐 𝑑 loc_{d} italic_l italic_o italic_c start_POSTSUBSCRIPT italic_d end_POSTSUBSCRIPT . If the robot has access to a spoon, the robot can use it to scoop the beans from the jar to the cooker. However, what if the robot did not have a spoon, but had a glass instead? By the definition of C S subscript 𝐶 𝑆 C_{S} italic_C start_POSTSUBSCRIPT italic_S end_POSTSUBSCRIPT , the agent is unaware that h ⁢ a ⁢ s ⁢ C ⁢ o ⁢ n ⁢ t ⁢ a ⁢ i ⁢ n ⁢ a ⁢ b ⁢ i ⁢ l ⁢ i ⁢ t ⁢ y ℎ 𝑎 𝑠 𝐶 𝑜 𝑛 𝑡 𝑎 𝑖 𝑛 𝑎 𝑏 𝑖 𝑙 𝑖 𝑡 𝑦 hasContainability italic_h italic_a italic_s italic_C italic_o italic_n italic_t italic_a italic_i italic_n italic_a italic_b italic_i italic_l italic_i italic_t italic_y (glass) is true, making the goal un-achievable. By our definition, creative problem solving is the process by which the agent uses some function f ⁢ ( ⋅ ) 𝑓 ⋅ f(\cdot) italic_f ( ⋅ ) to discover a new conceptual space: f ⁢ ( C S ) = C S ′ = C S ⁢ ∪ 𝑓 subscript 𝐶 𝑆 subscript superscript 𝐶 ′ 𝑆 subscript 𝐶 𝑆 f(C_{S})=C^{\prime}_{S}=C_{S}\mathop{\cup} italic_f ( italic_C start_POSTSUBSCRIPT italic_S end_POSTSUBSCRIPT ) = italic_C start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_S end_POSTSUBSCRIPT = italic_C start_POSTSUBSCRIPT italic_S end_POSTSUBSCRIPT ∪ { h ⁢ a ⁢ s ⁢ C ⁢ o ⁢ n ⁢ t ⁢ a ⁢ i ⁢ n ⁢ a ⁢ b ⁢ i ⁢ l ⁢ i ⁢ t ⁢ y ℎ 𝑎 𝑠 𝐶 𝑜 𝑛 𝑡 𝑎 𝑖 𝑛 𝑎 𝑏 𝑖 𝑙 𝑖 𝑡 𝑦 hasContainability italic_h italic_a italic_s italic_C italic_o italic_n italic_t italic_a italic_i italic_n italic_a italic_b italic_i italic_l italic_i italic_t italic_y  (glass)}. This would allow the agent to solve the previously unsolvable task by using the glass to scoop the beans instead.

Boden’s three forms of creativity denote three plausible functions for f ⁢ ( C X ) 𝑓 subscript 𝐶 𝑋 f(C_{X}) italic_f ( italic_C start_POSTSUBSCRIPT italic_X end_POSTSUBSCRIPT ) . CPS arises when the agent uses what it knows, to discover something new and the newly discovered knowledge is applied to solve a previously impossible task. We revisit the notion of conceptual spaces in Section 3.

In the remainder of this section, we discuss how typical task planning is achieved with LLVMs. We divide the discussion into three subsections based on the level of task planning abstraction where LLVMs are applied: a) high-level task planning, b) low-level task planning, and c) hybrid task planning. While not exhaustive, our review is meant to offer a general insight into how LLVMs are used for task planning, to identify entry points for introducing creative problem solving capabilities.

3.2 LLVMs for high-level task planning

Approaches for high-level task planning often involve using LLVMs to identify high-level goals for accomplishing a task. Some approaches to task planning with LLMs often take a user input specifying the task, and generate high-level task plans for accomplishing it. These approaches often use LLMs as a form of “knowledge base”, to extract actionable task plans from the models via appropriate prompting Huang et al. ( 2022 ) , further iterating over the generated task plan with repeated calls to the LLM as needed Prasad et al. ( 2023 ) .

In the context of Reinforcement Learning (RL), prior work has focused on using LLMs to suggest high-level goals for an RL agent Du et al. ( 2023 ) . Dubbed as ELLMs (Exploring with LLMs), an RL agent provides its current state to an LLM via a prompt, and receives a goal suggestion from the LLM that is then used to shape the reward and the agent exploration. Further work has extended this approach to incorporate the use of experience memory Zhang et al. ( 2023a ) . Existing approaches have also used LLMs to generate directed acyclic graphs composed of sub-goal states to aid the exploration of an RL agent Shukla et al. ( 2023 ) .

3.3 LLVMs for low-level task planning

Approaches for low-level task planning involve using LLMs to generate low-level code for performing a task. In contrast to high-level planning, where high-level goals and sub-goals are generated, these approaches use LLMs to directly generate low-level execution code via appropriate API calls Liang et al. ( 2023 ) . Other approaches have also investigated the capacity of LLMs to generate task plans via a low-level planning language such as PDDL Silver et al. ( 2023 ) , including iterating over the generated plan descriptions in case of errors Guan et al. ( 2023 ) . In terms of low-level planning using VLMs, prior work has introduced an approach that uses a diffusion model to generate robot trajectories conditioned on language and the current visual state of the robot Chen et al. ( 2023 ) .

3.4 Hybrid high and low-level planning with LLVMs

Hybrid approaches use LLVMs both for high-level goal generation as well as low-level planning. For instance, in Li et al. ( 2023 ) , user inputs are passed as LLM prompts to generate high-level plans. The high-level plans are then converted to low-level plans for robot execution via LLMs specialized for coding. Other approaches have used a high-level LLM planner, a VLM perceiver, and a low-level LLM planner for re-planning with both visual and language inputs Skreta et al. ( 2024 ) .

3.5 Summary

Given this overview, we see that LLVMs both at the high-level and low-level, can be modified to incorporate creative problem solving into task planning. For instance, the high-level task plans generated can encompass a novel substitution for a missing object, whereas the low-level task plan can generate an appropriate trajectory for creatively using the object. While the above approaches could, in principle, be studied within the framework of creative problem solving, that is not usually how the problem is formulated; there is a lack of paradigms for studying creative problem solving beyond just, “do you solve the problem or not?” . Creative problem solving needs a fundamental rethinking of the typical problem formulations and approaches in ML. The next section is aimed at ways in which ML approaches in LLVMs can be reformulated from the perspective of CC.

4 Augmenting LLVM embedding spaces for creative problem solving

In this section, we discuss how principles from CC can be extended to LLVMs for creative problem solving. We begin with Boden’s definition of “conceptual spaces” as “ [conceptual space] is the generative system that underlies the domain and defines a certain range of possibilities: chess moves, or molecular structures, or jazz melodies ” Boden ( 2005 ) , p.18 and “ … in short, any reasonably disciplined way of thinking ” Boden ( 1998 ) , p.214. By this definition, the embedding space of an LLVM describes its conceptual space or “ its way of thinking ”. Some evidence for this also comes from existing work that introduces an approach for enabling LLMs to interpret continuous embedding spaces via natural language. Given an embedding vector representing an interpolation of different concepts, the model is able to interpret a text prompt in the context of the supplied embedding Tennenholtz et al. ( 2023 ) . The embedding thus determines the model’s way of thinking. Hence, a discussion of enabling creative problem solving in LLVMs should target their embedding space. To this end, we explore two questions: a) how can LLVM embedding spaces be augmented to achieve creative problem solving, and b) what information should they be augmented with? Aligning with our original position, we show that CC literature can offer insights into these questions.

4.1 How can LLVM embedding spaces be augmented?

In this section, we draw parallels between Boden’s three forms of creativity and existing approaches in LLVMs. We further elaborate on how the three forms of creativity may enhance the potential of LLVMs to perform creative problem solving. We note that the ML approaches discussed in this section do not specifically perform creative problem solving. However, we discuss how they could potentially be extended to do so, by leveraging references from the CC literature.

4.1.1 Exploratory Creativity

Exploratory approaches involve exploration within the conceptual or equivalently, the embedding space of the model, and most closely relates to “search”. Note that the term “exploration” here differs from its usage in RL, instead referring to exploration through the model’s embedding space . Several existing approaches in the ML literature involve searching the output space of LLMs with the goal of improving the performance of these models. The “tree-of-thought” model generates a “tree” of next possible LLM outputs, and searches through the states via Breadth-first or Depth-first search to reach the desired goal state, often guided by heuristics Yao et al. ( 2023 ) . Numerous other approaches have built upon a similar strategy, such as using Monte-Carlo Tree Search (MCTS) Zhou et al. ( 2023 ); Feng et al. ( 2023 ) , beam search Zhang et al. ( 2023b ) or integrating pruning to remove sub-par candidates Golovneva et al. ( 2023 ) .

Extension of exploratory creativity to LLVMs: An important point to note here is that these approaches involve searching exclusively within the output “solution space” of the LLMs rather than directly operating in the embedding space itself. In contrast to operating in the solution space of the LLM, exploratory approaches directly within the LLMs’ embedding space would not be limited by what the LLM can generate as output – “ Some exploration merely shows us the nature of the relevant conceptual space that we had not explicitly noticed before ” Boden ( 2005 ) , p.18. To effectively reveal the full extent of the conceptual space for creative problem solving, the approach should not be limited by the outputs the LLVM can generate. Rather, the generated (creative) outputs itself should be the result of heuristic or non-heuristic based search within the model’s embedding space. However, to the best of our knowledge current approaches have not focused on LLVMs from this perspective, and have also not applied search to embedding spaces of Vision-LMs. Regardless, exploratory approaches are still limited by the dimensions of the model’s embedding space. “ To overcome a limitation in the conceptual space, one must change it in some way ” Boden ( 2005 ) , p.18 - this leads us to combinational and transformational creativity.

4.1.2 Combinational Creativity

Combinational approaches involve combining two concepts to create something new - “ A novel combination of two familiar ideas is something which did not happen before. ” Boden ( 1998 ) , p.213. We can broadly translate this to a function that takes in multiple concepts within an LLVM’s embedding space to output a novel concept.

One way of extending this definition to LLVMs involves applying cross-attention layers. The attention operation is defined as Vaswani et al. ( 2017 ) :

where, Q 𝑄 Q italic_Q , K 𝐾 K italic_K and V 𝑉 V italic_V denote query, keys and values respectively, and d k subscript 𝑑 𝑘 d_{k} italic_d start_POSTSUBSCRIPT italic_k end_POSTSUBSCRIPT denotes the dimensionality of the keys. Cross-attention involves passing K 𝐾 K italic_K and V 𝑉 V italic_V from a different model, e.g., in Flamingo Alayrac et al. ( 2022 ) , the keys and values represent visual input (from a separate vision encoder) and queries represent a language input. By applying cross attention in this manner, the embedding space of a model can be extended with capabilities of another model. In Bansal et al. ( 2024 ) the authors show that using cross-attention layers can help augment an anchor LLM with an augmenting LLM’s capabilities to perform a task that the anchor LLM was incapable of achieving before - hinting at some creative possibilities of this method.

Other approaches in LLVMs, while using “combinations” in some way, do not conform to the notion of combinational creativity . This includes, for instance, approaches that perform arithmetic combination of LLM weights to enhance the model performance Matena and Raffel ( 2022 ); Ilharco et al. ( 2022 ) . Or approaches that combine image and text embeddings via concatenation Kim et al. ( 2021 ) or a scaled dot product at the output Radford et al. ( 2021 ) . While these approaches may be useful in imparting multi-modal capabilities, however, they do not lead to combinational creativity since the combination occurs external to the models as opposed to within the model’s embedding space.

Extension of Combinational Creativity to LLVMs: The ML approaches described here involve combining embedding spaces across models. Existing approaches have not looked at combining concepts within the same model’s embedding space. The extension of combinational creativity to LLVMs is much more apparent in the sense of conceptual blending Fauconnier and Turner ( 2003 ) for generation of creative artifacts, e.g., via blending of artistic styles. However, the extension of combinational creativity to creative problem solving is less obvious, and CC literature offers us further insights for making this connection. Typical conceptual blending corresponds to a form of “aesthetic combination”, whereas creative problem solving would benefit from “functional combinations” Chen et al. ( 2018 ) . Functional combination combines the functions (as opposed to aesthetic) of two components, e.g., a coin combined with pliers could function as a makeshift screwdriver. The authors extend this framework to a combination of two nouns with a “base” noun (e.g., “pliers”) and “additive” noun (e.g., “coin”). An interesting possibility stems from this notion: Can a combination of embeddings of the same LLVM, corresponding to “base” and “additive” nouns (perhaps with some prior denoting the task), enable the LLVM to generate creative combinations of objects for solving a task? This question remains unexplored, and points to a potential research direction for LLVMs inspired by CC.

4.1.3 Transformational Creativity

Transformational approaches involve transforming existing conceptual spaces to produce new ones. Transforming conceptual spaces can involve “ altering existing rules ” Boden ( 1998 ) , p.216. One way of transforming a model’s embedding space involves fine-tuning or training Franceschelli and Musolesi ( 2023 ) . However, additional insight into transformational creative problem solving comes from prior work in CC, that describes creative problems as those with a poorly defined structure where a solution is not immediately apparent Olteteanu ( 2014 ) . And in such cases, “… re-representation being the process which transforms an ill-structured problem into a well-structured one with direct inference to a problem solution ” Olteteanu ( 2014 ) , p.1. The notion of “re-representing” or “redefining” the problem can be best captured in the input prompts provided to an LLVM. This most closely connects to prompt engineering and in-context learning (ICL).

Prompt engineering augments LLVMs with task specific hints, called prompts, to adapt the LLVM to new tasks Gu et al. ( 2023 ) . Relatedly, in-context learning is a prompting method that provides the LLVM with instructions for solving a new task without requiring additional training. Prior work has shown that in-context learning and gradient-based optimization are equivalent Von Oswald et al. ( 2023 ) , thus connecting ICL to training or fine-tuning.

Extension of transformational creativity to LLVMs: Task re-representations for creative problem solving, through prompting or ICL, has not been well explored within ML. Prompt engineering and ICL is a challenging task, since model performance depends strongly on the chosen prompts Rubin et al. ( 2021 ) , further compounded by the fact that creative problems are inherently poorly defined Olteteanu ( 2014 ) . However, useful insights can be derived from CC literature. For instance, regarding problems that require creatively re-purposing objects, the Object-replacement-object-composition (OROC) framework Olteţeanu and Falomir ( 2016 ) illustrates re-representations of tasks, that can be translated into prompts. The paper defines three different types of creative tasks involving objects, and their task re-representations as (from Olteţeanu and Falomir ( 2016 ) , p.16):

Replace an unfound object needed for a task with other objects present in the environment: “If I do not have an object X, which I would normally use because of its affordance 1 1 1 Affordance is defined as the relation between an agent, action and object, e.g., bowls have the “contain” affordance for humans. A ⁢ f X 𝐴 subscript 𝑓 𝑋 Af_{X} italic_A italic_f start_POSTSUBSCRIPT italic_X end_POSTSUBSCRIPT , what other object Y could I use, so that I can get a similar affordance, A ⁢ f X ≈ A ⁢ f Y 𝐴 subscript 𝑓 𝑋 𝐴 subscript 𝑓 𝑌 Af_{X}\approx Af_{Y} italic_A italic_f start_POSTSUBSCRIPT italic_X end_POSTSUBSCRIPT ≈ italic_A italic_f start_POSTSUBSCRIPT italic_Y end_POSTSUBSCRIPT ? ”

𝐴 subscript 𝑓 𝑌 1 𝐴 subscript 𝑓 𝑌 2 … 𝐴 subscript 𝑓 𝑌 𝑛 Af_{X}\approx Af_{X^{\prime}},Af_{X}\approx Af_{Y1}+Af_{Y2}+...+Af_{Yn} italic_A italic_f start_POSTSUBSCRIPT italic_X end_POSTSUBSCRIPT ≈ italic_A italic_f start_POSTSUBSCRIPT italic_X start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT end_POSTSUBSCRIPT , italic_A italic_f start_POSTSUBSCRIPT italic_X end_POSTSUBSCRIPT ≈ italic_A italic_f start_POSTSUBSCRIPT italic_Y 1 end_POSTSUBSCRIPT + italic_A italic_f start_POSTSUBSCRIPT italic_Y 2 end_POSTSUBSCRIPT + … + italic_A italic_f start_POSTSUBSCRIPT italic_Y italic_n end_POSTSUBSCRIPT ? ”

• subscript 𝑌 1 subscript 𝑌 2 … subscript 𝑌 𝑛 Y_{1};Y_{2};...;Y_{n} italic_Y start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ; italic_Y start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT ; … ; italic_Y start_POSTSUBSCRIPT italic_n end_POSTSUBSCRIPT which are components of object Y 𝑌 Y italic_Y could I use to obtain an object Y i ′ subscript superscript 𝑌 ′ 𝑖 Y^{\prime}_{i} italic_Y start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT with an equivalent or similar affordance, A ⁢ f X ≈ A ⁢ f Y ′ ⁢ i 𝐴 subscript 𝑓 𝑋 𝐴 subscript 𝑓 superscript 𝑌 ′ 𝑖 Af_{X}\approx Af_{Y^{\prime}i} italic_A italic_f start_POSTSUBSCRIPT italic_X end_POSTSUBSCRIPT ≈ italic_A italic_f start_POSTSUBSCRIPT italic_Y start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT italic_i end_POSTSUBSCRIPT ? ”

For task re-representation, affordances can refer to object properties that are relevant to the task, e.g., in some cases the shape may be relevant and in other cases, the material Olteţeanu and Falomir ( 2016 ) . Within LLVMs, the affordances A ⁢ f X 𝐴 subscript 𝑓 𝑋 Af_{X} italic_A italic_f start_POSTSUBSCRIPT italic_X end_POSTSUBSCRIPT or A ⁢ f Y 𝐴 subscript 𝑓 𝑌 Af_{Y} italic_A italic_f start_POSTSUBSCRIPT italic_Y end_POSTSUBSCRIPT can be defined via natural language, or other modalities such as images. In the following section, we present preliminary experiments on using LLVMs for object replacement, with prompts that are inspired by the above task re-representations. However, an in-depth application of these re-representations as defined in CC to in-context learning in LLVMs remains unexplored.

4.1.4 Summary

In the previous sections, we drew parallels between Boden’s three forms of creativity and approaches in LLVMs, further emphasizing how principles from CC can potentially help enable creative problem solving skills in these models.

Integration with task planning: Given the three methods, we see that transformational and combinational approaches may be especially aligned with LLVMs for high-level task planning. In contrast, exploratory methods may be suited to low-level planning, e.g., trajectory generation.

Creative problem solving as a combination of the three methods: An effective approach to creative problem solving may require all the three methods described in this section. While papers have explored chaining of LLMs within frameworks (often via prompts) Karpas et al. ( 2022 ); Ling et al. ( 2023 ) , the individual LLMs themselves do not exhibit the characteristics described here. Existing frameworks in CC have shown that achieving creative problem solving would take a combination of all three methods, each of which is triggered in different contexts Olteteanu ( 2014 ) . This presents potential opportunities for ML approaches that develop frameworks using multiple LLVMs, e.g., extending CC frameworks such as “ CreaCogs ” Olteţeanu and Falomir ( 2016 ) can be highly beneficial for productive developments in ML.

4.2 What information should LLVM embeddings be augemented with?

In the previous section, we discussed three methods for augmenting LLVM embedding spaces. In this section, we explore the question: “What information should be targeted by the three methods when augmenting the embedding space for creative problem solving?”. In the previous section, we discussed this in the context of OROC. According to the OROC framework Olteţeanu and Falomir ( 2016 ) , information about object affordances could enable models to re-represent the task, such that the solution becomes evident. We propose a small experiment to validate whether the principles of transformational creativity from OROC are useful to LLVMs. We note that creativity can occur in various contexts, e.g., creatively solving a math problem or creatively playing a chess move, each of which would require different information. However, to facilitate the discussion in this paper, we focus our scope on tasks that require innovatively replacing missing objects (OROC Task #1).

Note on embeddings vs. concepts: Our work connects “conceptual spaces” (or “concepts”) as defined in Computational Creativity literature, to “embedding spaces” (or “embeddings”) as defined in typical LM literature. We use “concepts” and “embeddings” interchangeably in this context. We make this connection to note that existing methods in Computational Creativity that operate on conceptual spaces translate to ML algorithms that operate on the LM’s embedding space. In this section, we connect the concept of “affordances” to the “embeddings” of the LLVMs in our experiments. Our goal is to show how the model can be prompted via an approach inspired by transformational creativity, to connect affordances of two seemingly distinct objects, e.g., a bowl and a spoon that appear distinct, but share the containability affordance.

4.2.1 Experiment Setup

We create a simple experiment setup that tests the “object replacement” principle from OROC, where we create test sets composed of images of objects for replacing one of five core objects: “Scoop”, “Hammer”, “Spatula”, “Toothpick”, and “Pliers”. We create two groups of tests: a) a nominal group where the actual object itself is available in each test set and requires no replacement (which serves as a form of baseline), and b) an object replacement group, where the nominal tool is missing and a creative replacement object should be chosen.

For each group, we create test sets with 4 objects each, chosen from a set of RGB images of 16 objects (Appendix Figure 3 ). We create 10 such test sets per core object (total 50 samples per model). Each test set only includes one ground truth object, along with three other random objects that will not suit as an appropriate replacement. In the nominal group, the ground truth is the actual object itself. In the object replacement group, the replacements are chosen based on self-assessment of the authors as (core object → absent → \xrightarrow{} start_ARROW start_OVERACCENT end_OVERACCENT → end_ARROW replacement): “Scoop” → absent → \xrightarrow{} start_ARROW start_OVERACCENT end_OVERACCENT → end_ARROW “Bowl”; “Hammer” → absent → \xrightarrow{} start_ARROW start_OVERACCENT end_OVERACCENT → end_ARROW “Saucepan”; “Spatula” → absent → \xrightarrow{} start_ARROW start_OVERACCENT end_OVERACCENT → end_ARROW “Knife”; “Toothpick” → absent → \xrightarrow{} start_ARROW start_OVERACCENT end_OVERACCENT → end_ARROW “Safety pin”; “Pliers” → absent → \xrightarrow{} start_ARROW start_OVERACCENT end_OVERACCENT → end_ARROW “Scissors”. For each test case, we pass the images in the test set along with a prompt. We record whether the ground truth object image was chosen by the model for the prompt (i.e., assigned highest output probability) 2 2 2 CLIP generates probabilities that given images correspond to a text. ViLT and LLaVA respond with a text, and we evaluate if the model responded “yes” with a high probability for the ground truth. .

The nominal group is subjected to one type of prompt: “ Can this object be used as a ⟨ c o r e _ o b j e c t ⟩ ? \bigl{\langle}core\_object\bigl{\rangle}? ⟨ italic_c italic_o italic_r italic_e _ italic_o italic_b italic_j italic_e italic_c italic_t ⟩ ? ”. In the object replacement group, each test case is subjected to four types of prompts:

Baseline (regular) prompt: Same prompt as used in the nominal cases to obtain a baseline.

Prompt prepended with affordance information: the prompt includes additional information about the desired object affordances specified as object features.

Prompt prepended with task information: the prompt includes additional information about the desired task.

Prompt prepended with task and affordance information: the prompt includes additional information on the task and object affordance.

Case #2 aligns with task re-representations of OROC, and we explore cases #3 and #4 for comparison. We formulate our affordance prompts as brief versions of OROC’s task re-representations. According to Olteţeanu and Falomir ( 2016 ) affordances can be defined using shape features, which we apply to the prompts here. The full set of prompts is shown in Appendix Table 2 . The models that we explore include versions of CLIP Radford et al. ( 2021 ) , LLaVA Liu et al. ( 2024 ) , and ViLT Kim et al. ( 2021 ) obtained from HuggingFace. We use different model sizes ( B ase, L arge, H uge) and patch sizes (14, 16, 32). The open-source code for reproducing our experiment results (including our dataset and test cases) is available at: https://github.com/lnairGT/creative-problem-solving-LLMs . Appendix C includes more details on the experiments.

4.2.2 Results

In Table 1 , we see the performances of the different models in the nominal test group, where the object requires no creative replacement. The models perform > 90 % absent percent 90 >90\% > 90 % in such cases (except for ViLT). In Figure 2 , we see the performances (accuracy shown on a 0.0 − 1.0 0.0 1.0 0.0-1.0 0.0 - 1.0 scale) of the models in the object replacement test cases, where the object requires a creative replacement. For reference, a model that randomly picks an object achieves about 30% overall accuracy. Figure 2 shows average accuracies for the different prompting strategies across random test sets. From Table 1 to Figure 2 (“regular”), the models perform poorly when they need to creatively reason about object replacements, highlighting their limitation. Comparing the “Regular” tab in Figure 2 to “Affordance”, we see a general improvement in model performances, when object affordance information is provided , consistent with description of the OROC framework Olteţeanu and Falomir ( 2016 ) . However, information about the task (Figure 2 , “Task” ) leads to mostly detrimental results. Information about task and affordances (Figure 2 , “Task + Affordance”) does not lead to substantial improvements either, and is also detrimental in certain cases. We note that there is quite a variance in performances across the different models, which may be partially attributed to the original training datasets of the models. These observations warrant further exploration beyond the scope of this paper. Appendix D includes a detailed, class-wise breakdown of the results.

4.2.3 Summary

While the experiments that we conducted are only preliminary, they offer some validity that the extension of principles in Computational Creativity can help overcome limitations of LLVMs in creative problem solving. The notion of task re-representation via improved prompting warrants further investigation in LLVMs, with regards to how the prompts can be generated automatically based on the creative task.

The models used in our experiments have all been trained jointly in visual and text domains. Multi-modal prompting capabilities may be useful for achieving creative problem solving. It can be quite challenging to describe affordances in words (example of “hammers” in our tests) and they may be better described through other means, e.g., images or depth maps or spectral data for material properties Erickson et al. ( 2020 ) . This would require application of multi-modal LLVMs that can process a variety of data types Girdhar et al. ( 2023 ); Han et al. ( 2023 ) . Computational creativity can offer insights into meaningful representations of these different modalities that would help achieve creative problem solving, e.g., whether object material or shape matters more for one task vs. another Olteţeanu and Falomir ( 2016 ) .

It is also worth noting that the creative problem solving examples in our experiments are human-centric. For instance, robots may not have similar capabilities as humans to manipulate bowls for scooping. In such cases, LLVMs need to account for the affordances as described with respect to the agent , in order to derive creative solutions. However, that adds another level of complexity, yet to be explored, since these models are typically trained on human-centric data.

5 Evaluation of Creativity

An important discussion in the context of creative problem solving is, how can creative problem solving be evaluated? . Prior work has proposed that creativity necessitates both novelty and value Boden ( 1998 ); Runco and Jaeger ( 2012 ) , where the former guarantees that the generated outputs of a creative process are original, and the latter ensures that the generated outputs are useful. In the context of CPS, novelty refers to the discovery of new concepts (as defined in section 3.1 ), whereas value insists that the newly discovered concepts successfully solve the task. Hence, benchmarks for CPS should specifically evaluate how the task was solved (novelty and value) rather than the typical ML evaluation of whether the task was successful or not (value only). Some existing approaches that make this distinction describe problem settings that can be used to measure CPS skills of LLMs through the implicit integration of novelty and value measurements Tian et al. ( 2023 ); Naeini et al. ( 2023 ); Bisk et al. ( 2020 ); Talmor et al. ( 2022 ) . In Tian et al. ( 2023 ) , the authors create a dataset of 1600 real-world problems that necessarily involve creative reasoning abilities. Their proposed benchmark involves identifying novel approaches that can accomplish the given task (value). Similarly, in Naeini et al. ( 2023 ) , the authors introduce the Only-Connect-Wall (OCW) dataset to measure CPS capabilities of LLMs. The authors in Bisk et al. ( 2020 ) explore physical commonsense reasoning that is more generally applicable, beyond object-based creative problems. The authors introduce Physical Interaction: Question Answering, or PIQA consisting of 16,000 QA pairs where each question is paired with two possible common-sense solutions with a ground truth. In Talmor et al. ( 2022 ) , the authors introduce CommonSenseQA 2.0 (CSQA2) dataset consisting of both object-based and non-object based creative problems. The dataset consists of 14,343 questions distributed across 1,868 distinct topics. Currently, to the best of our knowledge, there are no standard benchmarks available to measure CPS skills of VLMs, although our preliminary experiments show one way to measure this using the task of object substitution.

6 Conclusion and Future Work

In this paper, we argued that an effective approach for enabling creative problem solving – currently a key limitation of LLVMs – should derive from Computational Creativity literature. To emphasize this at each juncture, we discussed the specific principles from CC that can be extended to achieve creative problem solving in LLVMs, describing the potential for further research with these insights. It is rare to see special tracks or workshops targeted at Computational Creativity within more prestigious ML conferences. These programs typically focus on creative artifact generation and art (such as the NeurIPS Workshop on Machine Learning for Creativity and Design NeurIPS ( 2022 ) or the recent tutorial at EMNLP on Creative Natural Language Generation Chakrabarty et al. ( 2023 ) ), but do not discuss CPS, thus failing to bridge the gap between CC and ML. We hope to see a deeper integration of the CC communities at such strong ML venues. We hope to encourage the reader to view creative problem solving and ML holistically, through the lens of Computational Creativity.

7 Limitations

Literature outside of Computational Creativity that enables CPS is unexplored: Our paper predominantly focuses on CC literature. This work does not cover literature beyond CC that can potentially inform creative problem solving in LLVMs. Although CC literature broadly encompasses psychology, neuroscience and philosophy, our future work seeks to explore specific literature within these sub-domains and discuss their applicability to creative problem solving and ML.

Lack of an explicit creative problem solving algorithm for LLVMs: Since the scope of our work aligns with a position paper, we have not focused on developing a concrete algorithm for creative problem solving in LLVMs. The prompting strategies explored in our preliminary experiments are manually specified, and our work does not elaborate on how these prompts may be automatically discovered. While our paper seeks to address some of the key gaps that prevent the application of CC literature to ML, there are still several unanswered questions when it comes to the practical implementation of an ML approach: e.g., what is a good representation for concepts that facilitate creative problem solving (symbolic, non-symbolic, or hybrid)? What is a good problem formulation for a given creative problem solving task (planning or learning)? etc. However, these questions are not directly answered within the scope of our work.

8 Ethical Considerations

The authors do not have specific ethical considerations to be highlighted with respect to this work.

• Alayrac et al. (2022) Jean-Baptiste Alayrac, Jeff Donahue, Pauline Luc, Antoine Miech, Iain Barr, Yana Hasson, Karel Lenc, Arthur Mensch, Katherine Millican, Malcolm Reynolds, et al. 2022. Flamingo: a visual language model for few-shot learning. Advances in Neural Information Processing Systems , 35:23716–23736.
• Anna Tong and Hu (2023) Jeffrey Dastin Anna Tong and Krystal Hu. 2023. Openai researchers warned board of ai breakthrough ahead of ceo ouster, sources say. https://www.reuters.com/technology/sam-altmans-ouster-openai-was-precipitated-by-letter-board-about-ai-breakthrough-2023-11-22/. [Online; accessed 19-Jan-2024].
• Atkeson et al. (2018) Christopher G Atkeson, PW Babu Benzun, Nandan Banerjee, Dmitry Berenson, Christoper P Bove, Xiongyi Cui, Mathew DeDonato, Ruixiang Du, Siyuan Feng, Perry Franklin, et al. 2018. What happened at the darpa robotics challenge finals. The DARPA robotics challenge finals: Humanoid robots to the rescue , pages 667–684.
• Bansal et al. (2024) Rachit Bansal, Bidisha Samanta, Siddharth Dalmia, Nitish Gupta, Shikhar Vashishth, Sriram Ganapathy, Abhishek Bapna, Prateek Jain, and Partha Talukdar. 2024. Llm augmented llms: Expanding capabilities through composition. arXiv preprint arXiv:2401.02412 .
• BBC (2012) BBC. 2012. Us navy funds ’macgyver’ robot that can create tools. https://www.bbc.com/news/technology-19902954 . [Online; accessed 9-April-2024].
• Bisk et al. (2020) Yonatan Bisk, Rowan Zellers, Jianfeng Gao, Yejin Choi, et al. 2020. Piqa: Reasoning about physical commonsense in natural language. In Proceedings of the AAAI conference on artificial intelligence , volume 34, pages 7432–7439.
• Boden (1998) Margaret A. Boden. 1998. Creativity and Artificial Intelligence. Artificial Intelligence , 1-2:347–356.
• Boden (2005) Margaret A. Boden. 2005. What is creativity? Creativity in human evolution and prehistory .
• Bubeck et al. (2023) Sébastien Bubeck, Varun Chandrasekaran, Ronen Eldan, Johannes Gehrke, Eric Horvitz, Ece Kamar, Peter Lee, Yin Tat Lee, Yuanzhi Li, Scott Lundberg, et al. 2023. Sparks of artificial general intelligence: Early experiments with gpt-4. arXiv preprint arXiv:2303.12712 .
• Cass (2005) Stephen Cass. 2005. Apollo 13, we have a solution. IEEE Spectrum On-line, 04 , 1.
• Chakrabarty et al. (2023) Tuhin Chakrabarty, Vishakh Padmakumar, He He, and Nanyun Peng. 2023. Creative natural language generation. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing: Tutorial Abstracts , pages 34–40.
• Chen et al. (2023) Lili Chen, Shikhar Bahl, and Deepak Pathak. 2023. Playfusion: Skill acquisition via diffusion from language-annotated play. In Conference on Robot Learning , pages 2012–2029. PMLR.
• Chen et al. (2018) Liuqing Chen, Pan Wang, Feng Shi, Ji Han, Peter Childs, et al. 2018. A computational approach for combinational creativity in design. In DS 92: Proceedings of the DESIGN 2018 15th International Design Conference , pages 1815–1824.
• Copet et al. (2023) Jade Copet, Felix Kreuk, Itai Gat, Tal Remez, David Kant, Gabriel Synnaeve, Yossi Adi, and Alexandre Défossez. 2023. Simple and controllable music generation. arXiv preprint arXiv:2306.05284 .
• DeCaro et al. (2011) Marci S DeCaro, Robin D Thomas, Neil B Albert, and Sian L Beilock. 2011. Choking under pressure: multiple routes to skill failure. Journal of experimental psychology: general , 140(3):390.
• Du et al. (2023) Yuqing Du, Olivia Watkins, Zihan Wang, Cédric Colas, Trevor Darrell, Pieter Abbeel, Abhishek Gupta, and Jacob Andreas. 2023. Guiding pretraining in reinforcement learning with large language models. arXiv preprint arXiv:2302.06692 .
• Erickson et al. (2020) Zackory Erickson, Eliot Xing, Bharat Srirangam, Sonia Chernova, and Charles C Kemp. 2020. Multimodal material classification for robots using spectroscopy and high resolution texture imaging. In 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) , pages 10452–10459. IEEE.
• Fauconnier and Turner (2003) Gilles Fauconnier and Mark Turner. 2003. Conceptual blending, form and meaning. Recherches en communication , 19:57–86.
• Fei et al. (2022) Nanyi Fei, Zhiwu Lu, Yizhao Gao, Guoxing Yang, Yuqi Huo, Jingyuan Wen, Haoyu Lu, Ruihua Song, Xin Gao, Tao Xiang, et al. 2022. Towards artificial general intelligence via a multimodal foundation model. Nature Communications , 13(1):3094.
• Feng et al. (2023) Xidong Feng, Ziyu Wan, Muning Wen, Ying Wen, Weinan Zhang, and Jun Wang. 2023. Alphazero-like tree-search can guide large language model decoding and training. arXiv preprint arXiv:2309.17179 .
• Franceschelli and Musolesi (2023) Giorgio Franceschelli and Mirco Musolesi. 2023. On the creativity of large language models. arXiv preprint arXiv:2304.00008 .
• Gilhooly (2016) Kenneth J Gilhooly. 2016. Incubation and intuition in creative problem solving. Frontiers in psychology , 7:1076.
• Girdhar et al. (2023) Rohit Girdhar, Alaaeldin El-Nouby, Zhuang Liu, Mannat Singh, Kalyan Vasudev Alwala, Armand Joulin, and Ishan Misra. 2023. Imagebind: One embedding space to bind them all .
• Gizzi et al. (2022) Evana Gizzi, Lakshmi Nair, Sonia Chernova, and Jivko Sinapov. 2022. Creative problem solving in artificially intelligent agents: A survey and framework. Journal of Artificial Intelligence Research , 75:857–911.
• Goertzel (2014) Ben Goertzel. 2014. Artificial general intelligence: concept, state of the art, and future prospects. Journal of Artificial General Intelligence , 5(1):1.
• Golovneva et al. (2023) O. Golovneva, S. O’Brien, R. Pasunuru, T. Wang, L. Zettlemoyer, M. Fazel-Zarandi, and A. Celikyilmaz. 2023. Pathfinder: Guided search over multi-step reasoning paths. arXiv preprint arXiv:2312.05180 .
• Grudin and Jacques (2019) Jonathan Grudin and Richard Jacques. 2019. Chatbots, humbots, and the quest for artificial general intelligence. In Proceedings of the 2019 CHI conference on human factors in computing systems , pages 1–11.
• Gu et al. (2023) Jindong Gu, Zhen Han, Shuo Chen, Ahmad Beirami, Bailan He, Gengyuan Zhang, Ruotong Liao, Yao Qin, Volker Tresp, and Philip Torr. 2023. A systematic survey of prompt engineering on vision-language foundation models. arXiv preprint arXiv:2307.12980 .
• Guan et al. (2023) Lin Guan, Karthik Valmeekam, Sarath Sreedharan, and Subbarao Kambhampati. 2023. Leveraging pre-trained large language models to construct and utilize world models for model-based task planning. arXiv preprint arXiv:2305.14909 .
• Guilford (1967) Joy P Guilford. 1967. Creativity: Yesterday, today and tomorrow. The Journal of Creative Behavior , 1(1):3–14.
• Hammond et al. (2013) Adam Hammond, Julian Brooke, and Graeme Hirst. 2013. A tale of two cultures: Bringing literary analysis and computational linguistics together. In Proceedings of the Workshop on Computational Linguistics for Literature , pages 1–8.
• Han et al. (2023) Jiaming Han, Kaixiong Gong, Yiyuan Zhang, Jiaqi Wang, Kaipeng Zhang, Dahua Lin, Yu Qiao, Peng Gao, and Xiangyu Yue. 2023. Onellm: One framework to align all modalities with language .
• Hélie and Sun (2010) Sebastien Hélie and Ron Sun. 2010. Incubation, insight, and creative problem solving: a unified theory and a connectionist model. Psychological review , 117(3):994.
• Huang et al. (2022) Wenlong Huang, Pieter Abbeel, Deepak Pathak, and Igor Mordatch. 2022. Language models as zero-shot planners: Extracting actionable knowledge for embodied agents. In International Conference on Machine Learning , pages 9118–9147. PMLR.
• Ilharco et al. (2022) Gabriel Ilharco, Marco Tulio Ribeiro, Mitchell Wortsman, Suchin Gururangan, Ludwig Schmidt, Hannaneh Hajishirzi, and Ali Farhadi. 2022. Editing models with task arithmetic. arXiv preprint arXiv:2212.04089 .
• Karpas et al. (2022) Ehud Karpas, Omri Abend, Yonatan Belinkov, Barak Lenz, Opher Lieber, Nir Ratner, Yoav Shoham, Hofit Bata, Yoav Levine, Kevin Leyton-Brown, Dor Muhlgay, Noam Rozen, Erez Schwartz, Gal Shachaf, Shai Shalev-Shwartz, Amnon Shashua, and Moshe Tenenholtz. 2022. Mrkl systems: A modular, neuro-symbolic architecture that combines large language models, external knowledge sources and discrete reasoning .
• Kim et al. (2021) Wonjae Kim, Bokyung Son, and Ildoo Kim. 2021. Vilt: Vision-and-language transformer without convolution or region supervision. In International Conference on Machine Learning , pages 5583–5594. PMLR.
• Lahikainen et al. (2024) Joonas Lahikainen, Nadia M Ady, and Christian Guckelsberger. 2024. Creativity and markov decision processes. arXiv preprint arXiv:2405.14966 .
• Li et al. (2023) Boyi Li, Philipp Wu, Pieter Abbeel, and Jitendra Malik. 2023. Interactive task planning with language models. arXiv preprint arXiv:2310.10645 .
• Liang et al. (2023) Jacky Liang, Wenlong Huang, Fei Xia, Peng Xu, Karol Hausman, Brian Ichter, Pete Florence, and Andy Zeng. 2023. Code as policies: Language model programs for embodied control. In 2023 IEEE International Conference on Robotics and Automation (ICRA) , pages 9493–9500. IEEE.
• Ling et al. (2023) Zhan Ling, Yunhao Fang, Xuanlin Li, Tongzhou Mu, Mingu Lee, Reza Pourreza, Roland Memisevic, and Hao Su. 2023. Unleashing the creative mind: Language model as hierarchical policy for improved exploration on challenging problem solving .
• Liu et al. (2024) Haotian Liu, Chunyuan Li, Qingyang Wu, and Yong Jae Lee. 2024. Visual instruction tuning. Advances in neural information processing systems , 36.
• Ma et al. (2023) Yuxi Ma, Chi Zhang, and Song-Chun Zhu. 2023. Brain in a vat: On missing pieces towards artificial general intelligence in large language models. arXiv preprint arXiv:2307.03762 .
• Matena and Raffel (2022) Michael S Matena and Colin A Raffel. 2022. Merging models with fisher-weighted averaging. Advances in Neural Information Processing Systems , 35:17703–17716.
• Moor et al. (2023) Michael Moor, Oishi Banerjee, Zahra Shakeri Hossein Abad, Harlan M Krumholz, Jure Leskovec, Eric J Topol, and Pranav Rajpurkar. 2023. Foundation models for generalist medical artificial intelligence. Nature , 616(7956):259–265.
• Moruzzi (2020) Caterina Moruzzi. 2020. Artificial creativity and general intelligence. Journal of Science and Technology of the Arts .
• Naeini et al. (2023) Saeid Naeini, Raeid Saqur, Mozhgan Saeidi, John Giorgi, and Babak Taati. 2023. Large language models are fixated by red herrings: Exploring creative problem solving and einstellung effect using the only connect wall dataset. arXiv preprint arXiv:2306.11167 .
• NeurIPS (2022) NeurIPS. 2022. Workshop on machine learning for creativity and design. https://nips.cc/virtual/2022/workshop/49965. [Online; accessed 19-Jan-2024].
• Olteteanu (2014) Ana-Maria Olteteanu. 2014. Two general classes in creative problem-solving? an account based on the cognitive processess involved in the problem structure-representation structure relationship. Publications of the Institute of Cognitive Science .
• Olteţeanu and Falomir (2016) Ana-Maria Olteţeanu and Zoe Falomir. 2016. Object replacement and object composition in a creative cognitive system. towards a computational solver of the alternative uses test. Cognitive Systems Research , 39:15–32.
• Patel (2023) Dwarkesh Patel. 2023. Llms need search for problem solving - shane legg (deepmind founder). https://www.youtube.com/watch?v=qulfo6-54k0. [Online; accessed 19-Jan-2024].
• Pennachin and Goertzel (2007) Cassio Pennachin and Ben Goertzel. 2007. Contemporary approaches to artificial general intelligence. In Artificial general intelligence , pages 1–30. Springer.
• Prasad et al. (2023) Archiki Prasad, Alexander Koller, Mareike Hartmann, Peter Clark, Ashish Sabharwal, Mohit Bansal, and Tushar Khot. 2023. Adapt: As-needed decomposition and planning with language models. arXiv preprint arXiv:2311.05772 .
• Radford et al. (2021) Alec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh, Gabriel Goh, Sandhini Agarwal, Girish Sastry, Amanda Askell, Pamela Mishkin, Jack Clark, et al. 2021. Learning transferable visual models from natural language supervision. In International conference on machine learning , pages 8748–8763. PMLR.
• Rombach et al. (2021) Robin Rombach, Andreas Blattmann, Dominik Lorenz, Patrick Esser, and Björn Ommer. 2021. High-resolution image synthesis with latent diffusion models. 2022 ieee. In CVF Conference on Computer Vision and Pattern Recognition (CVPR) , pages 10674–10685.
• Rubin et al. (2021) Ohad Rubin, Jonathan Herzig, and Jonathan Berant. 2021. Learning to retrieve prompts for in-context learning. arXiv preprint arXiv:2112.08633 .
• Runco and Jaeger (2012) Mark A Runco and Garrett J Jaeger. 2012. The standard definition of creativity. Creativity research journal , 24(1):92–96.
• Sarathy and Scheutz (2018) Vasanth Sarathy and Matthias Scheutz. 2018. Macgyver problems: Ai challenges for testing resourcefulness and creativity. Advances in Cognitive Systems , 6:31–44.
• Shevlin et al. (2019) Henry Shevlin, Karina Vold, Matthew Crosby, and Marta Halina. 2019. The limits of machine intelligence: Despite progress in machine intelligence, artificial general intelligence is still a major challenge. EMBO reports , 20(10):e49177.
• Shukla et al. (2023) Yash Shukla, Wenchang Gao, Vasanth Sarathy, Alvaro Velasquez, Robert Wright, and Jivko Sinapov. 2023. Lgts: Dynamic task sampling using llm-generated sub-goals for reinforcement learning agents. arXiv preprint arXiv:2310.09454 .
• Silver et al. (2023) Tom Silver, Soham Dan, Kavitha Srinivas, Joshua B Tenenbaum, Leslie Pack Kaelbling, and Michael Katz. 2023. Generalized planning in pddl domains with pretrained large language models. arXiv preprint arXiv:2305.11014 .
• Skreta et al. (2024) Marta Skreta, Zihan Zhou, Jia Lin Yuan, Kourosh Darvish, Alán Aspuru-Guzik, and Animesh Garg. 2024. Replan: Robotic replanning with perception and language models. arXiv preprint arXiv:2401.04157 .
• Talmor et al. (2022) Alon Talmor, Ori Yoran, Ronan Le Bras, Chandra Bhagavatula, Yoav Goldberg, Yejin Choi, and Jonathan Berant. 2022. Commonsenseqa 2.0: Exposing the limits of ai through gamification. arXiv preprint arXiv:2201.05320 .
• Tennenholtz et al. (2023) Guy Tennenholtz, Yinlam Chow, Chih-Wei Hsu, Jihwan Jeong, Lior Shani, Azamat Tulepbergenov, Deepak Ramachandran, Martin Mladenov, and Craig Boutilier. 2023. Demystifying embedding spaces using large language models. arXiv preprint arXiv:2310.04475 .
• Tian et al. (2023) Yufei Tian, Abhilasha Ravichander, Lianhui Qin, Ronan Le Bras, Raja Marjieh, Nanyun Peng, Yejin Choi, Thomas L Griffiths, and Faeze Brahman. 2023. Macgyver: Are large language models creative problem solvers? arXiv preprint arXiv:2311.09682 .
• Turner et al. (2020) MC Turner, LV Duggan, BA Glezerson, and SD Marshall. 2020. Thinking outside the (acrylic) box: a framework for the local use of custom-made medical devices. Anaesthesia .
• Van Heerden and Bas (2021) Imke Van Heerden and Anil Bas. 2021. Ai as author–bridging the gap between machine learning and literary theory. Journal of Artificial Intelligence Research , 71:175–189.
• Vaswani et al. (2017) Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N Gomez, Łukasz Kaiser, and Illia Polosukhin. 2017. Attention is all you need. Advances in neural information processing systems , 30.
• Ventura (2014) Dan Ventura. 2014. Can a computer be lucky? and other ridiculous questions posed by computational creativity. In Artificial General Intelligence: 7th International Conference, AGI 2014, Quebec City, QC, Canada, August 1-4, 2014. Proceedings 7 , pages 208–217. Springer.
• Von Oswald et al. (2023) Johannes Von Oswald, Eyvind Niklasson, Ettore Randazzo, João Sacramento, Alexander Mordvintsev, Andrey Zhmoginov, and Max Vladymyrov. 2023. Transformers learn in-context by gradient descent. In International Conference on Machine Learning , pages 35151–35174. PMLR.
• Wallas (1926) Graham Wallas. 1926. The art of thought . 24. Harcourt, Brace.
• Wang (2023) Brian Wang. 2023. Openai q* could be based upon a* search without expansions. https://www.nextbigfuture.com/2023/11/openai-q-could-be-based-upon-a-search-without-expansions.html. [Online; accessed 19-Jan-2024].
• Wiggins (2006) Geraint A Wiggins. 2006. A preliminary framework for description, analysis and comparison of creative systems. Knowledge-based systems , 19(7):449–458.
• Xi et al. (2023) Zhiheng Xi, Wenxiang Chen, Xin Guo, Wei He, Yiwen Ding, Boyang Hong, Ming Zhang, Junzhe Wang, Senjie Jin, Enyu Zhou, et al. 2023. The rise and potential of large language model based agents: A survey. arXiv preprint arXiv:2309.07864 .
• Yao et al. (2023) Shunyu Yao, Dian Yu, Jeffrey Zhao, Izhak Shafran, Thomas L Griffiths, Yuan Cao, and Karthik Narasimhan. 2023. Tree of thoughts: Deliberate problem solving with large language models. arXiv preprint arXiv:2305.10601 .
• Zhang et al. (2023a) Danyang Zhang, Lu Chen, Situo Zhang, Hongshen Xu, Zihan Zhao, and Kai Yu. 2023a. Large language model is semi-parametric reinforcement learning agent. arXiv preprint arXiv:2306.07929 .
• Zhang et al. (2023b) Shun Zhang, Zhenfang Chen, Yikang Shen, Mingyu Ding, Joshua B Tenenbaum, and Chuang Gan. 2023b. Planning with large language models for code generation. arXiv preprint arXiv:2303.05510 .
• Zhou et al. (2023) Andy Zhou, Kai Yan, Michal Shlapentokh-Rothman, Haohan Wang, and Yu-Xiong Wang. 2023. Language agent tree search unifies reasoning acting and planning in language models. arXiv preprint arXiv:2310.04406 .

Appendix A Alternate Definitions of Creative Problem Solving

Prior work by Olteţeanu Olteteanu ( 2014 ) defines CPS from an object affordance perspective, where affordances broadly refer to action possibilities for objects, e.g., cups are pour-able and doors are open-able. The authors in Olteteanu ( 2014 ) define creative problems as nominal problem solving tasks that have a poor representational structure, and as “ the ability of a cognitive, natural, or artificial system to use new objects to solve a problem, other than the ones that have been stored in its memory as tools for that specific purpose (if any), or to create those objects by putting together objects or parts of objects the system has access to. Depending on the problem, objects can be either physical or abstract/informational (concepts, problem templates, heuristics or other forms of representations) ”. However, this definition is primarily object-creativity centered, and does not cover a wider range of creative problems.

Follow-up work by Sarathy and Scheutz Sarathy and Scheutz ( 2018 ) , define “ Macgyver-esque ” creativity as a planning task that involves “ generating, executing, and learning strategies for identifying and solving seemingly unsolvable real-world problems ”. They introduce the “ MacGyver Problem ” (MGP) as a planning problem with an unreachable goal state. Through the modification of the agent’s domain knowledge (through domain expansion and domain contraction ), the agent must discover new information and incorporate it into its existing domain knowledge, allowing the agent to accomplish the task. The domain expansion and contraction processes align with the divergent-convergent model of creative problem solving Guilford ( 1967 ) . The definition of an MGP aligns well with the formulation of planning problems in ML, but less with learning or hybrid planning-learning approaches.

Appendix B Alternate theories on creative problem solving and their applications to ML

While there is exhaustive literature regarding theories on general creativity, we focus specifically on creative problem solving, with three well received works: Divergent-Convergent Thinking Guilford ( 1967 ) , Explicit-Implicit Interaction Theory Hélie and Sun ( 2010 ) , and the Creative Systems Framework Wiggins ( 2006 ) . We discuss their applicability to ML in addition to the literature discussed in the main body of this paper. Our goal in this section is to further widen the discussion on integrating CC and ML to achieve creative problem solving in LLVMs, with additional literature.

B.0.1 Divergent-Convergent Thinking

In Guilford ( 1967 ) , the authors discuss the notion of “divergent-convergent” thinking. Divergent thinking or “divergent-production” (DP) abilities involve a more open-ended generation of a variety of ideas, whereas convergent thinking focuses on applying specific ideas to solve the problem.

Applicability to CPS in LLVMs: Prior work by Tian et al. ( 2023 ) have demonstrated the applicability of “divergent-convergent” thinking towards solving Macgyver problems. Similar in spirit to our experiments with VLMs in Section 4.2.1 , the authors prompt LLMs with descriptions of objects to enable the LLMs to reason about solving the task. Their work is, to the best of our knowledge, the only direct example demonstrating the value of CC literature in enabling CPS in LLMs.

B.0.2 Explicit-Implicit Interaction Theory

In Hélie and Sun ( 2010 ) , the authors introduce the Explicit-Implicit Interaction (EII) theory, building upon the seminal work in Wallas ( 1926 ) , that describes four stages of creativity: Preparation, incubation, illumination (i.e., insight), and verification. Preparation refers to the initial stage of searching in many different directions, which may fail to find a solution (i.e., impasse) in case of ill-defined problems (as is the case with CPS). Following an impasse, the incubation phase begins, where attention is not devoted to solving the problem. Over a period of time, illumination is the manifestation of the solution to the problem within the conscious thought (i.e., “Aha” moment). Finally, verification involves using deliberative thinking to assess if the solution indeed solves the problem.

Applicability to CPS in LLVMs: The authors in Hélie and Sun ( 2010 ) incorporate the four stages via a concrete computational method into the CLARION cognitive architecture. Prior work has also introduced a CPS framework for ML approaches inspired by the four stages Gizzi et al. ( 2022 ) . In their work, “preparation” aligns with problem formulation, either task learning or planning. Incubation and illumination aligns with knowledge representation (symbolic, non-symbolic, or hybrid), and knowledge manipulation (functions that manipulate the conceptual space). Lastly, verification aligns with evaluation (via simulation, real-world platforms, or benchmarks). Although these works do not explicitly cover LLVMs and related algorithms, they demonstrate the value of integrating CC literature in ML, and can serve as useful starting points for ML approaches towards creative problem solving in LLVMs.

B.0.3 Creative Systems Framework

In Wiggins ( 2006 ) , the author expands on Boden’s levels further in the context of a framework that formalizes creative systems. The paper defines: a) creative system, b) creative behavior, c) novelty, and d) value. The paper also discusses formalized notion of a universe of possibilities , and conceptual spaces . Crucially, the work describes the characteristics of a creative agent, that can help distinguish modes of failures within a creative system, namely: a) hopeless uninspiration – where there are no valued concepts within the universe; b) conceptual uninspiration – where there are no valued concepts within the conceptual space of the agent; and c) generative uninspiration – where an agent is unable to find a valued concept owing to the specific method (e.g., search) employed.

Applicability to CPS in LLVMs: While the discussion of novelty, value and conceptual spaces in Wiggins ( 2006 ) aligns with our descriptions in Section 4 , the different modes of uninspiration offers potential ways to assess failure modes in LLVMs. This allows agents to distinguish between systems where creative problem solving is not possible (hopeless uninspiration), as compared to systems where the conceptual space or the methodology for searching the conceptual space, may be at fault (conceptual or generative uninspiration). Although this approach has not been expanded in existing literature, it presents a promising direction for an evaluation framework that can distinguish CPS from non-CPS problems.

B.1 A potential link between creative problem solving and general intelligence

Existing literature hints at a potential link between creative problem solving and Artificial General Intelligence (AGI) - systems that are broadly capable of solving almost all tasks that humans can Shevlin et al. ( 2019 ) . For instance, in Moruzzi ( 2020 ) , p.85., the author argues that there exists a strong correlation between creativity and AGI: “ … features that systems need to develop in order to achieve general intelligence are aspects that they need to possess also to earn the attribute creative ”. In Goertzel ( 2014 ) , the author compiles a list of competencies deemed essential for achieving AGI, including creative capacities like “ conceptual invention ” and “ creative constructive play with objects ”. The processes of “insight” or “incubation” often associated with creative problem solving Hélie and Sun ( 2010 ); Gilhooly ( 2016 ) is also considered important for AGI Ventura ( 2014 ) . Taken together, it is likely that any promising vision of AGI would be incomplete without creative problem solving .

Alongside the heavy ongoing discussion of AGI surrounding LLVMs Bubeck et al. ( 2023 ); Fei et al. ( 2022 ); Ma et al. ( 2023 ); Xi et al. ( 2023 ); Moor et al. ( 2023 ); Grudin and Jacques ( 2019 ) , there is often little to no discussion of creative problem solving or Computational Creativity within mainstream ML. As described in Moruzzi ( 2020 ) , p.96, “ The investigation on the nature of creativity and on how it manifests itself not only in human but also in animal and artificial systems should, thus, not be intended as a niche discussion but, rather, as a fundamental research which can lay the foundations for further studies in artificial intelligence and its relation to humans ”. We hope that this work will encourage discussions of creative problem solving and Computational Creativity alongside discussions on AGI.

Appendix C Experiment Settings

Prompt type

Prompt

Regular

“can this object be used as a scoop?” “can this object be used as a hammer?” “can this object be used as a spatula?” “can this object be used as a toothpick?” “can this object be used as pliers?” “scoops must be concave and hollow. can this object be used as a scoop?” “hammers must be heavy and have a handle attached to a cylinder at the end. can this object be used as a hammer?” “spatulas must have a handle attached to a flat surface at the end. can this object be used as a spatula?” “toothpicks must have a pointed tip. can this object be used as a toothpick?” “pliers must have two-prongs. can this object be used as pliers?” “scoops can transfer beans from one jar to another jar. can this object be used as a scoop?” “hammers can hit a nail into the wall. can this object be used as a hammer?” “spatulas can spread butter onto a pan. can this object be used as a spatula?” “toothpicks can pick food caught between the teeth. can this object be used as a toothpick?” “pliers can grab a coin. can this object be used as pliers?” “scoops can transfer beans from one jar to another jar. scoops are concave and hollow. can this object be used as a scoop?” “hammers can hit a nail into the wall. hammers have a handle attached to a cylinder at the end. can this object be used as a hammer?” “spatulas can spread butter onto a pan. spatulas have a handle attached to a flat surface at the end. can this object be used as a spatula?” “toothpicks can pick food caught between the teeth. toothpicks have a pointed tip. can this object be used as a toothpick?” “pliers can grab a coin. pliers have two-prongs. can this object be used as pliers?” C.1 Data: Test images Figure 3 shows the test set of 16 RGB images of objects used for the object substitution task. From the shown image dataset, we create test sets with 4 objects each, chosen from the set of 16 object images. We create 10 such test sets per core object (total 50 samples per model). Each test set only includes one ground truth object, along with three other random objects that will not suit as an appropriate replacement. In the nominal group, the ground truth is the actual object itself. In the object replacement group, the replacements are chosen based on self-assessment of the authors as (core object → absent → \xrightarrow{} start_ARROW start_OVERACCENT end_OVERACCENT → end_ARROW replacement): “Scoop” → absent → \xrightarrow{} start_ARROW start_OVERACCENT end_OVERACCENT → end_ARROW “Bowl”; “Hammer” → absent → \xrightarrow{} start_ARROW start_OVERACCENT end_OVERACCENT → end_ARROW “Saucepan”; “Spatula” → absent → \xrightarrow{} start_ARROW start_OVERACCENT end_OVERACCENT → end_ARROW “Knife”; “Toothpick” → absent → \xrightarrow{} start_ARROW start_OVERACCENT end_OVERACCENT → end_ARROW “Safety pin”; “Pliers” → absent → \xrightarrow{} start_ARROW start_OVERACCENT end_OVERACCENT → end_ARROW “Scissors”. C.2 Model: Checkpoints For all the models, we use pre-trained HuggingFace checkpoints, with no additional training or fine-tuning. The models are of different architecture sizes and patch sizes: “CLIP-B-32” uses the “openai/clip-vit-base-patch32” which is a base model with a patch size of 32. “CLIP-B-16” uses “openai/clip-vit-base-patch16” – a base model with patch size of 16. “CLIP-L-14” uses “openai/clip-vit-large-patch14” – a large model with patch size of 14. “CLIP-H-14” uses “laion/CLIP-ViT-H-14-laion2B-s32B-b79K” which is a “huge” model, with a patch size of 14. This model is trained with the 2 billion sample English subset of LAION-5B. For LLaVA, we use the “llava-hf/llava-1.5-7b-hf” with 7B parameters, version 1.5. Lastly, “VILT-B-32” uses “dandelin/vilt-b32-finetuned-vqa” trained for visual question answering. However, there is limited data available on HuggingFace regarding the model. C.3 Prompts used in testing In this section, we discuss the prompts used in the different testing conditions (see Table 2 ). We explore four classes of prompts for the creative object substitution task: “Regular”, “Affordance”, “Task” and “Task and affordance”. Regular prompts involve a direct prompt as to whether a given object will suffice as a substitute for the missing object. Affordance prompts, adds information about the desired affordances that are essential for replacing the missing object. Task prompts adds additional information on the task to be performed as context for whether a given object can be used as replacement for the missing object. Lastly, task and affordance prompts combine the task and object affordance information within the prompt. C.4 Testing Procedure For each test case, we pass the images in the test set along with a prompt belonging to one of the four classes described in Table 2 . We record whether the ground truth object image was chosen by the model for the prompt (i.e., assigned highest output probability). CLIP generates probabilities that given images correspond to a text. ViLT responds with a text, and we evaluate if the model responded “yes” with a high probability for the ground truth. C.5 Testing Infrastructure We used NVIDIA-A100 GPUs to run the evaluation. However, the models are not too large and we have tested and confirmed that the code can be executed on CPU only as well. Appendix D Continued Experiment Results In this section, we show the class-wise breakdown of the different models for the different prompting strategies (Figures 4 - 7 ). We note that “hammers” present a particularly challenging case for all the models, perhaps due to the fact that correlating affordance of a hammer to a saucepan textually is difficult. In contrast, all models with the augmented prompts typically perform well in the case of creatively replacing “toothpick” with “safety pin” – presumably indicating that specifying the relevant affordance textually in this case provides sufficient information. We repeated each experiment across multiple random seeds and found similar performances, showing that our general findings hold across different random cases. Generally, specifying object affordance information in the prompts leads to improved model performance. Module 5: Thinking and Analysis Solving problems creatively, learning outcomes. Describe the role of creative thinking skills in problem-solving Problem-Solving with Creative Thinking Creative problem-solving is a type of problem-solving. It involves searching for new and novel solutions to problems. Unlike critical thinking, which scrutinizes assumptions and uses reasoning, creative thinking is about generating alternative ideas—practices and solutions that are unique and effective. It’s about facing sometimes muddy and unclear problems and seeing how things can be done differently—how new solutions can be imagined. [1] You have to remain open-minded, focus on your organizational skills, and learn to communicate your ideas well when you are using creative thinking to solve problems. If an employee at a café you own suggests serving breakfast in addition to the already-served lunch and dinner, keeping an open mind means thinking through the benefits of this new plan (e.g., potential new customers and increased profits) instead of merely focusing on the possible drawbacks (e.g., possible scheduling problems, added start-up costs, loss of lunch business). Implementing this plan would mean a new structure for buying, workers’ schedules and pay, and advertising, so you would have to organize all these component areas. And finally, you would need to communicate your ideas on how to make this new plan work not only to the staff who will work the new shift, but also to the public who frequent your café and the others you want to encourage to try your new hours. We need creative solutions throughout the workplace—whether board room, emergency room, or classroom. It was no fluke that the 2001 revised Bloom’s cognitive taxonomy, originally developed in 1948, placed a new word at the apex— creating . That  creating is the highest level of thinking skills. Bloom’s Taxonomy is an important learning theory used by psychologists, cognitive scientists, and educators to demonstrate levels of thinking. Many assessments and lessons you’ve seen during your schooling have likely been arranged with Bloom’s in mind. Researchers recently revised it to place creativity—invention—as the highest level “Because we’ve always done it that way” is not a valid reason to not try a new approach. It may very well be that the old process is a very good way to do things, but it also may just be that the old, comfortable routine is not as effective and efficient as a new process could be. The good news is that we can always improve upon our problem-solving and creative-thinking skills—even if we don’t consider ourselves to be artists or creative. The following information may surprise and encourage you! Creative thinking (a companion to critical thinking) is an invaluable skill for college students. It’s important because it helps you look at problems and situations from a fresh perspective. Creative thinking is a way to develop novel or unorthodox solutions that do not depend wholly on past or current solutions. It’s a way of employing strategies to clear your mind so that your thoughts and ideas can transcend what appear to be the limitations of a problem. Creative thinking is a way of moving beyond barriers. [2] As a creative thinker, you are curious, optimistic, and imaginative. You see problems as interesting opportunities, and you challenge assumptions and suspend judgment. You don’t give up easily. You work hard. [3] Is this you? Even if you don’t yet see yourself as a competent creative thinker or problem-solver, you can learn solid skills and techniques to help you become one. Creative Problem-Solving: Fiction and Facts As you continue to develop your creative thinking skills, be alert to perceptions about creative thinking that could slow down progress. Remember that creative thinking and problem-solving are ways to transcend the limitations of a problem and see past barriers. It’s a way to think outside the box. Creative Problem-Solving: Fiction and Facts FICTION FACTS 1 Every problem has only one solution (or one right answer). The goal of problem-solving is to solve the problem, and most problems can be solved in any number of ways. If you discover a solution that works, it’s a good solution. Other people may think up solutions that differ from yours, but that doesn’t make your solution wrong or unimportant. What is the solution to “putting words on paper”? Fountain pen, ballpoint, pencil, marker, typewriter, printer, printing press, word-processing . . .? 2 The best answer or solution or method has already been discovered. Look at the history of any solution and you’ll see that improvements, new solutions, and new right answers are always being found. What is the solution to human transportation? The ox or horse, the cart, the wagon, the train, the car, the airplane, the jet, or the space shuttle? What is the best and last? 3 Creative answers are technologically complex. Only a few problems require complex technological solutions. Most problems you’ll encounter need only a thoughtful solution involving personal action and perhaps a few simple tools. Even many problems that seem to require technology can be addressed in other ways. 4 Ideas either come or they don’t. Nothing will help— certainly not structure. There are many successful techniques for generating ideas. One important technique is to include structure. Create guidelines, limiting parameters, and concrete goals for yourself that stimulate and shape your creativity. This strategy can help you get past the intimidation of the blank page. For example, if you want to write a story about a person who gained insight through experience, you can stoke your creativity by limiting or narrowing your theme to “a young girl in Cambodia who escaped the Khmer Rouge to find a new life as a nurse in France.” Apply this type of specificity and structure to any creative endeavor. creative problem-solving: a practice that seeks new and novel solutions to problems, often by using imagination rather than linear reason "Critical and Creative Thinking, MA." University of Massachusetts Boston . 2016. Web. 16 Feb 2016. ↵ Mumaw, Stefan. "Born This Way: Is Creativity Innate or Learned?" Peachpit. Pearson, 27 Dec 2012. Web. 16 Feb 2016. ↵ Harris, Robert. "Introduction to Creative Thinking." Virtual Salt. 2 Apr 2012. Web. 16 Feb 2016. ↵ Ibid. ↵ College Success. Authored by : Linda Bruce. Provided by : Lumen Learning. License : CC BY: Attribution College Success. Authored by : Amy Baldwin. Provided by : OpenStax. Located at : https://openstax.org/books/college-success/pages/7-2-creative-thinking . License : CC BY: Attribution Text adaptation. Authored by : Claire. Provided by : Ivy Tech. Located at : http://ivytech.edu/ . License : CC BY: Attribution 303 IMAGES What Is Critical Thinking And Creative Problem Solving What Is Critical Thinking And Creative Problem Solving : Critical Creative and Critical Thinking= Problem Solving (Adapted from What Is Creative Problem-Solving and How to Master It with These 8 Creative thinking and problem Creative Thinking Lesson 2 Problem solving and Decision making COMMENTS Critical Thinking vs. Problem-Solving: What's the Difference? Critical thinking vs. problem-solving Critical thinking and problem-solving can both help you resolve challenges, but the two practices have distinct purposes and strategies. Here are some differences between the two skills: Critical thinking This is a mode of thinking, compared to problem-solving, which is a set of solution-oriented strategies. What Is Creative Problem-Solving & Why Is It Important? Creative problem-solving primarily operates in the ideate phase of design thinking but can be applied to others. This is because design thinking is an iterative process that moves between the stages as ideas are generated and pursued. This is normal and encouraged, as innovation requires exploring multiple ideas. Critical Thinking vs. Creative Thinking Critical Thinking vs. Creative Thinking Creative thinking is a way of looking at problems or situations from a fresh perspective to conceive of something new or original. Critical thinking is the logical, sequential disciplined process of rationalizing, analyzing, evaluating, and interpreting information to make informed judgments and/or decisions. Exploring the Difference: Creative Thinking vs. Critical Thinking Creative thinking and critical thinking are two essential cognitive skills that play a significant role in problem-solving, decision-making, and innovation. While creative thinking involves generating new ideas, thinking outside the box, and exploring different perspectives , critical thinking focuses on analyzing, evaluating, and questioning ... Creative Thinking vs. Critical Thinking It emphasizes logical reasoning, evidence-based thinking, and the ability to identify biases and fallacies. While creative thinking focuses on generating ideas, critical thinking focuses on evaluating and refining those ideas. Both thinking processes are essential for problem-solving, decision-making, and personal growth. What Is Creative Thinking? Definition and Examples 1. Put Yourself in a Box. Creative thinking is about "thinking outside the box," but putting limitations on your problem-solving can help you think more freely and innovatively. For example, if someone tells you to make dinner, you may struggle to come up with a meal you don't always cook. Solving Problems Creatively Problem-Solving with Creative Thinking. Creative problem-solving is a type of problem-solving. It involves searching for new and novel solutions to problems. Unlike critical thinking, which scrutinizes assumptions and uses reasoning, creative thinking is about generating alternative ideas—practices and solutions that are unique and effective. Creative Problem Solving Creative problem solving (CPS) is a way of solving problems or identifying opportunities when conventional thinking has failed. It encourages you to find fresh perspectives and come up with innovative solutions, so that you can formulate a plan to overcome obstacles and reach your goals. In this article, we'll explore what CPS is, and we'll ... The science behind creativity Specifically, creativity often involves coordination between the cognitive control network, which is involved in executive functions such as planning and problem-solving, and the default mode network, which is most active during mind-wandering or daydreaming (Beaty, R. E., et al., Cerebral Cortex, Vol. 31, No. 10, 2021). Understanding Creativity DO IT and Min Basadur's Simplex embed the two approaches within problem-solving processes. While these may be considered 'overkill' when dealing with minor problems, they provide excellent frameworks for solving difficult and serious ones. The Creative Frame of Mind. Often the only difference between creative and uncreative people is self ... What is Creative Problem Solving? Creative thinking and problem solving are core parts of user experience (UX) design. Note: the abbreviation "CPS" can also refer to cyber-physical systems. Creative problem solving might sound somewhat generic or broad. ... It's important to know the difference between the two styles of thinking and when to practice them. This is why in a ... The art of solving problems: Comparing the similarities and differences creative problem solving methods - Lateral Thinking, Synectics and Creative Problem Solving (CPS). Specifically, the intent was to compare the three methods against focused dimensions of analysis. These dimensions of analysis were philosophical, theoretical, structural, functional and efficacy. Specific criteria were generated for each ... Critical thinking vs Creative thinking Both critical thinking and creative thinking are used for solving problems, only in different ways. For critical thinking, the process is structured and methodical. For creative thinking, the process is fluid and somewhat experimental. Both thinking strategies are useful, with neither being innately superior to the other and in some unexpected ... Creative Thinking vs. Critical Thinking: Unleashing the Power of Both Creative thinking and critical thinking are two essential cognitive skills that play a crucial role in problem-solving, decision-making, and innovation. While creative thinking involves generating new ideas, thinking outside the box, and exploring unconventional solutions, critical thinking focuses on analyzing, evaluating, and making logical ... Design Thinking vs Problem-Solving: Understanding the Differences Design thinking is more about creating new solutions, problem solving is more about finding solutions to existing problems. Design thinking starts by understanding the user's needs and then using that understanding to create new and innovative solutions. Problem-solving, on the other hand, is focused on finding a solution to a specific problem. What Is The Difference Between Critical Thinking And Creative Thinking While critical thinking involves the systematic evaluation of information and arguments, creative thinking is focused on generating novel and innovative ideas and solutions. Balancing these thinking styles results in enhanced productivity, better communication, and more creative and effective problem-solving. What Is Design Thinking & Why Is It Important? The first, and arguably most important, step of design thinking is building empathy with users. By understanding the person affected by a problem, you can find a more impactful solution. On top of empathy, design thinking is centered on observing product interaction, drawing conclusions based on research, and ensuring the user remains the focus ... Thinking Critically and Creatively Critical and creative thinking skills are perhaps the most fundamental skills involved in making judgments and solving problems. They are some of the most important skills I have ever developed. I use them everyday and continue to work to improve them both. The ability to think critically about a matter—to analyze a question, situation, or ... Thinking skills Divergent thinking: Breaking a topic apart to explore its various components and then generating new ideas and solutions. Critical Thinking: Analysis and evaluation of information, beliefs, or knowledge. Creative Thinking: Generation of new ideas breaking from established thoughts, theories, rules, and procedures. Critical Thinking versus Problem Solving The first step to enhancing your critical thinking and problem solving skills is to think about them, become aware of them, then you can actively practice to improve them. Critical thinking and problem-solving are two important "soft" or essential skills hiring managers are looking for. According to a Linkedin survey, 57% of business ... What Is the Difference Between Analytical and Creative Problem Solving Analytical and creative problem solving abilities rely on different skill sets. Analytical thinking is also referred to as logical thinking, while creative thinking can also be called "lateral" thinking. Sometimes the difference is described in terms of left-brain, or analytical, and right brain, or ... Critical Thinking vs Analytical Thinking vs Creative Thinking Creative thinking is imagining your own idea of the perfect cookie and then making it a reality for others to enjoy. So to put it technically, and in a way less likely to induce cravings: Analytical thinking is the act of breaking down complex pieces of information into smaller and more understandable components or principles (Thaneerananon, et ... Full article: Cultivating Critical Thinking Skills: a Pedagogical Study Kong et al. (2014) found that problem-based learning helps nursing students to improve their critical thinking skills. Faculty who intentionally integrated lessons involving writing responses and problem-solving found measurable critical thinking gains among students in geography and social studies courses (Hew & Cheung, 2014; Sziarto et al, 2014). Creative Problem Solving in Large Language and Vision Models While recent state-of-the-art large language models (LLMs) and vision-language models (VLMs) have demonstrated competency in artistic endeavours Rombach et al. (); Copet et al. (), creative problem solving continues to be a shortcoming of these models (we use LLVM to denote the umbrella of both LLMs and VLMs). For instance, in Bubeck et al. (), the authors point out that "discontinuous tasks ... Solving Problems Creatively Creative problem-solving is a type of problem-solving. It involves searching for new and novel solutions to problems. Unlike critical thinking, which scrutinizes assumptions and uses reasoning, creative thinking is about generating alternative ideas—practices and solutions that are unique and effective. It's about facing sometimes muddy and ... essay homework paper phd project resume review study thesis