Of all the forms of human intelligence that one would expect an AI to emulate, few would place creativity at the top of the list. Creativity is deeply mysterious—and frustratingly fleeting. It defines us as human beings, and seems to defy the cold logic behind the silicon curtain of machines.
However, the use of artificial intelligence in creative activities is growing.
New artificial intelligence tools such as DALL-E and Midjourney are increasingly being part of creative production, some of which have already begun Awarded for its creative achievements. Its social and economic impact is growing – to name just one example, the potential of AI to generate new creative content is a decisive tipping point behind AI. Hollywood screenwriters strike.
If our recent research The amazing originality of artificial intelligence By any indication, the emergence of AI-based creativity — and examples of its promise and dangers — may well be just getting started.
Combination of novelty and practicality
When people are at their most creative, they respond to a need, goal, or problem by creating something new—a product or solution that didn’t exist before.
In this sense, creativity is the act of combining existing resources (ideas, materials, knowledge) in novel ways that are useful or satisfying. Many times, the results of creative thinking are also surprising, leading to things that the creator didn’t—perhaps couldn’t—foresee.
It might involve an invention, an unexpected joke or a groundbreaking theory of physics. It could be a unique arrangement of notes, rhythms, sounds and lyrics that results in a new song.
So, as a researcher of creative thinking, I immediately noticed something interesting about the content generated by the latest versions of artificial intelligence (including GPT-4).
When prompted for tasks requiring creative thinking, the novelty and usefulness of the GPT-4 output reminded me of the creative ideas submitted by students and colleagues I’ve worked with as teachers and entrepreneurs.
The ideas are different and surprising, yet related and useful. And, when needed, be very imaginative.
Consider the following prompt given to GPT-4: “Suppose that one day of the week all children turned into giants. What would happen?” GPT-4 generates ideas about culture, economics, psychology, politics, interpersonal communication , transportation, entertainment, and more – many ideas are surprising and unique in terms of the novel connections that are made.
This combination of novelty and utility is hard to come by, as most scientists, artists, writers, musicians, poets, chefs, founders, engineers and scholars can attest.
Yet AI seems to be doing just that, and doing it well.
test artificial intelligence
Join Creativity and Entrepreneurship Researchers Christian Burge and christian guildI decided to test the creativity of an AI by having it take the Torrance Test of Creative Thinking, or TTCT.
TTCT Prompts Candidates to Participate The creativity required for real-life tasks: Ask questions, how to be more resourceful or efficient, speculate about cause and effect, or improve a product. As the example above shows, it might ask candidates to suggest improvements to children’s toys or to imagine the consequences of a hypothetical situation.
These tests are not designed to measure historical creativity, a term used by some researchers to describe the transformative brilliance of figures such as Mozart and Einstein.Instead, it assesses an individual’s general creativity, often referred to as mental or personal creativity.
In addition to running TTCT eight times through GPT-4, we also tested it on 24 undergraduate students.
All results are assessed by trained examiners from the Scholastic Examination Service, a private testing company that provides scores for the TTCT. They didn’t know in advance that some of the tests they were going to score were done by AI.
Because Academic Testing Service is a private company, it does not share its prompts with the public. This ensures that GPT-4 cannot crawl past prompts and their responses on the internet. In addition, the company maintains a database of thousands of tests completed by college students and adults, providing a large additional control group against which to compare AI scores.
our result?
GPT-4 scored in the top 1% of test-takers for the originality of their ideas. Based on our research, we believe this marks one of the first examples of AI matching or surpassing the original human capacity for thought.
In short, we believe that AI models like GPT-4 can generate ideas that people find unexpected, novel, and unique.Other researchers have come to similar conclusions Their research on artificial intelligence and creativity.
Yes, creativity can be assessed
The emerging creativity of artificial intelligence is astonishing for a number of reasons.
On the one hand, many people outside the research community still believe in creativity undefined, let alone score. Yet for thousands of years, products of human novelty and ingenuity have been prized and bought and sold. Creative work has been defined and scored in fields such as psychology since at least the 1950s.
Journalistic Models of People, Products, Processes, and CreativityThe concept of “creativity,” coined by researcher Mel Rhodes in 1961, attempts to categorize the various ways in which creativity has been understood and assessed to date. Since then, people’s understanding of creativity has continued to deepen.
Still others were surprised that the term “creativity” might apply to non-human entities such as computers. On this point, we tend to agree with cognitive scientist Margaret Boden, who sees the question as to whether the term creativity should be applied to artificial intelligence. Philosophical rather than scientific.
AI’s founder foresaw its creativity
It is worth noting that we only looked at the output of AI in our study.we didn’t study its creative process, which can be very different from the human thought process or the environment in which an idea arises. If we define creativity as requiring humans, then by definition we have to conclude that AI cannot be creative.
But whatever the debate over the definition of creativity and the creative process, the latest versions of AI-generated products are novel and useful. We believe this satisfies the definition of creativity that currently dominates the fields of psychology and science.
Furthermore, the creativity of the current iteration of AI is not entirely unexpected.
In their now famous proposal 1956 Dartmouth Summer Research Project on Artificial IntelligenceThe founders of artificial intelligence emphasized their desire to simulate “all aspects of learning or any other characteristic of intelligence” — including creativity.
In the same proposal, computer scientist Nathaniel Rochester reveals his motive: “How can I build a machine that exhibits ingenuity in solving problems?”
Apparently, the founders of artificial intelligence believed that creativity, including originality of ideas, is one of the specific forms of human intelligence that machines can imitate.
To me, the astonishing creativity scores of GPT-4 and other AI models highlight a more pressing problem: There are, to date, very few official programs and programs dedicated to human creativity and ability in American schools. Course implementation. nurture its development.
In this sense, the creative capabilities now enabled by artificial intelligence may provide “sputnik moment“For educators and others interested in enhancing human creativity, including those who see creativity as a fundamental condition for personal, social, and economic growth.
Eric GuzicAssistant Clinical Professor of Management, University of Montana
This article is reproduced from dialogue Licensed under Creative Commons.read source article.
Svlook
