August 20, 2024 By Sascha Brodsky 3 min read

As artificial intelligence continues to reshape industries, its impact on creativity has been a hot topic of debate. Can machines truly be creative? Or will they simply mimic human ingenuity? A new study published in Nature Human Behaviour sheds light on this question, suggesting that generative AI might be more of a creative collaborator than a replacement.

Researchers put ChatGPT, the popular AI chatbot, through its paces, pitting it against Google searches and unaided human thinking. The result? Large language models (LLMs) came out on top, helping people generate more creative ideas across a range of tasks.

In one experiment, participants were challenged to design a new toy using everyday items like paper clips and water bottles. The LLM-assisted group’s ideas sparkled, rated 10% more creative by judges than those relying on Google.

But here’s the kicker: ChatGPT held its own even in tasks thought to need a human touch. When asked to repurpose sentimental items, the AI-assisted responses still scored higher on the creativity scale.

AI vs. human creativity: The debate

Not everyone’s ready to crown ChatGPT the new king of creativity. Jacob Murel, Senior Technical Content Creator at IBM and former writing instructor, offers a reality check: “I don’t think it would ever replace humans,” Murel argues. “It can only produce what’s in its training data, so it can’t create a truly original idea.”

Sabrina Habib, Associate Professor and Visual Communications Sequence Head at the University of South Carolina, shares a more nuanced perspective. “I research how technology impacts creativity, and I learned that while AI tools like ChatGPT can enhance ideas, they can also hinder the human creative process,” Habib explains. She emphasizes the importance of human-first ideation: “When we generate ideas without technology, we engage in critical thinking and imagination, which are key for personal growth and innovation.”

Habib’s research suggests a balanced approach to AI-assisted creativity. “My suggestion is to always brainstorm without technology first, then include AI in the process,” she advises. This method preserves the benefits of human intuition while leveraging AI’s strengths.

The potential reshaping of creative industries is a crucial concern for Habib. “I think that many creative roles will shift to curating and improving AI-generated content, increasing efficiency and speed, but also raising questions of authorship,” she notes. This shift could have far-reaching implications for various sectors.

Murel sees potential in AI as a tool for inspiration and creativity. His students found LLMs helpful in unexpected ways: “They would give chatbot prompts, and it would then help them think about their reading and writing in a new way,” he explains.

Balancing AI and human creativity

So, what’s ChatGPT’s secret sauce? The study suggests it’s great at mixing and matching diverse concepts into coherent, articulate responses. Think of it as a high-speed idea blender capable of making connections that might take humans much longer to discover.

But don’t retire your thinking cap just yet. Murel advises using AI as a springboard, not a crutch. “Use it to help brainstorm ideas,” he suggests. “Have it write you a first draft, and then edit and rewrite what it says, because it will give you mediocre prose.”

Habib warns of potential pitfalls in over-relying on AI. “One of the consequences will be the production of homogeneous ideas, as my colleagues and I found in one of our studies,” she cautions. “Trust will be an issue as AI doesn’t always make ethical choices and often lacks context and care.”

For businesses, AI could mean new ways to spark innovation and streamline product development. AI could help teams quickly generate and evaluate a wide range of ideas, allowing human creators to focus their energy on refining and implementing the most promising concepts.

However, Habib emphasizes the enduring value of human creativity: “Human creativity is vital for cultural relevance and unique perspectives, which have always been key to driving innovation and will remain critical over time.”

Report: Forrester Total Economic Impact study of watsonx Assistant
Was this article helpful?
YesNo

More from Artificial intelligence

Examining synthetic data: The promise, risks and realities

3 min read - As artificial intelligence reshapes industries worldwide, developers are grappling with an unexpected challenge: a shortage of high-quality, real-world data to train their increasingly sophisticated models. Now, a potential solution is emerging from an unlikely source—data that doesn't exist in reality at all. Synthetic data, artificially generated information designed to mimic real-world scenarios, is rapidly gaining traction in AI development. It promises to overcome data bottlenecks, address privacy concerns, and reduce costs. However, as the field evolves, questions about its limitations…

Advance your enterprise Journey to Hybrid Cloud and AI powered by AIOps on Z

2 min read - Thanks to rising costs, skills shortages and ever-growing security threats, businesses must adapt quickly to shifts in demand patterns brought on by a digital workforce and rapidly changing buyer behavior. That requires putting extra emphasis on the resiliency and performance of your business processes and supporting applications. For larger IT organizations with increasingly hybrid and complex application landscapes that often include IBM Z®, it’s essential to take a comprehensive approach to IT operations. The challenge becomes “How do you effectively sift through terabytes of…

What’s new in watsonx Code Assistant for Z 2.1

2 min read - Enterprises today are adopting generative AI faster than before. According to an IBM survey, “83% of C-suite executives say modernizing applications is central to their organization’s business strategy. And 89% say that using generative AI in application modernization projects will drive growth by improving existing products and services and building new capabilities.” IBM® watsonx Code Assistant™ for Z is built to assist developers, architects and system programmers in modernizing their mainframe applications through the use of generative AI and automation.…

IBM Newsletters

Get our newsletters and topic updates that deliver the latest thought leadership and insights on emerging trends.
Subscribe now More newsletters