They studied 370,000 US college essays and discovered AI’s effects on creativity

I am a big fan of technology. I’ve blissfully given over my spatial reasoning to Google Maps. I use artificial intelligence (AI) to chase down articles, do research, fix my grammar mistakes and whip up last-minute school-night recipes.

But I’ve recently drawn a sharp line in the sand: No AI for writing.

I’m not talking about expense reports or routine e-mails. I mean actual writing, and the creative brainstorming that precedes it to explore different perspectives or develop novel insights.

Increasingly, many people I talk to – from students to teachers to peers – tell me that they think it is okay to use AI chatbots for brainstorming as long as they do the “real work” of writing.

But this misunderstands something critical. Brainstorming is the work that is fundamental to writing. As a researcher studying AI’s effects on education, I have concluded that these tools only superficially improve writing. The bigger and more alarming impact they have is to constrict our full range of thoughts and our ability to generate original and useful ideas – what we call creative thinking.

This seems to be especially true for students. AI’s smooth sentences, elegant transitions and rich vocabulary give the illusion of expansive creativity and individuality. But the underlying ideas often converge into a few homogenised categories.

The erosion of creative thinking means young people will struggle to navigate uncertainty. Workers will strain to adapt to a shifting labour market. And society will miss out on the new ideas that can solve complex problems and enhance lives.

For the past eight years, Georgetown University neuroscientist Adam Green has been leading a national research team tracking the range of novel ideas that college-bound high school students present in their application essays, before and after the introduction of ChatGPT.

In one study, he and his team examined personal statements from more than 370,000 students, and found that after ChatGPT became available, their essays suddenly used diverse and colourful language, but lacked truly creative ideas. And the linguistic cover-up worked. Post-ChatGPT essays were rated as more “creative” by human judges, even if the substance of the essays trod familiar territory.

In a separate study, the team found that human-written essays offered up to eight times more new ideas than those produced by AI.

Another experiment run by a different research team compared short stories written by humans to those written with AI assistance.

As with the student essays in Green’s study, AI-assisted essays had more interesting vocabulary and were rated more enjoyable to read, but the underlying story lines were more homogenous. Distinctive and offbeat ideas – with surprising characters or unusual settings – are often shunted to the side when AI is involved.

We now have a technology that can generate words separately from the thoughts they represent. When a chatbot writes, it is predicting the next word that is most likely to make a “good” sentence or essay, based on the text it’s been trained on. It can identify sophisticated and creative word patterns independently of whether the underlying ideas represent something new.

When teenagers write essays, the work reflects their thoughts and personalities, their attempts to make meaning of their experiences. When we search for words, we are sifting through the same brain networks that form connections between ideas.

A student who writes, “I’ll always think of learning to swim when I see a kite flying”, is connecting unique personal experiences in her life, which until recently, is a clear signal of truly creative thinking.

Another way AI interaction can narrow ideas is through the power of suggestion.

Once a chatbot suggests a direction, humans tend to lock in on it. The conversational nature of AI can make it difficult to distinguish where the user’s thinking ends and the bot’s begins, making it effortless for people to adopt AI-generated perspectives as their own.

It is easy to see how an impressionable teenager could forgo writing the unconventional essay – about, say, what it feels like to play jazz or cook with your grandmother – in favour of whatever AI suggests.

Even more problematic, Green’s research shows that AI has the largest homogenising impact on students who are farthest from the mean and have unique perspectives, including neurodivergent students and those from racial and linguistic minorities.

This is not to say that AI can never support human creativity. Workers with deep knowledge of their craft can use AI to streamline technical or administrative tasks in order to focus on the parts of their jobs where originality lives – including teachers having more time to devise engaging lessons and illustrators devoting more attention to developing visual concepts.

AI gives specialists the time they need to do what humans do best: brainstorming ideas to creatively solve problems.

Our species’ ability to come up with unexpected and original ideas is something to protect and nurture. That is especially true for today’s adolescents. A world where creative thinking flourishes is a world that has a better chance to weather the changes that AI will bring. NYTIMES