AI has become good at producing polished writing, answering questions, and generating ideas in seconds.
But John Nosta, an innovation theorist and founder of the think tank Nostalab, worries the technology could gradually crowd out the rare, original thinking that has historically driven humanity’s biggest breakthroughs.
Nosta compares human thinking to a bell curve. Most people cluster around the middle, he said, while the people responsible for the biggest breakthroughs sit at the extremes, or what he called the “tails.”
“What we’re seeing is that large language models, particularly when they train themselves on what’s called recursive data — that is, data generated by AI itself — that bell curve starts to change,” Nosta told Business Insider.
“We lose the tail. It just becomes one big bump,” he said. “What happens there is that it sort of blends into a sea of mediocrity.”
“Einstein was a tail. Newton was a tail. Picasso was a tail,” he added.
‘Society depends on those outliers’
For Nosta, innovation comes from unusual people willing to think differently.
“Where’s the outlier? Where’s the weird idea? Where’s the person who sees something nobody else sees? That’s what concerns me,” he said of AI.
He said recursive AI systems can gravitate toward consensus rather than originality, making it harder for unconventional thinking to emerge.
“Those strange ideas get pushed toward the middle because the computational dynamics of the AI push those strange outliers away,” he said.
Nosta doesn’t think AI will eliminate creativity altogether. He thinks it could make originality rarer by nudging people toward the same patterns of thinking.
“Society depends on those outliers,” he said.
The evidence so far is mixed. Some research suggests relying heavily on AI can steer users toward more conventional, homogeneous ideas, while other studies have found the technology can enhance creativity when people use it to explore possibilities rather than replace their own thinking.
Chris Best, the CEO of newsletter platform Substack, has said that while AI could flood the internet with “AI slop,” it could also usher in “a future where there’s way more creative leverage” by giving independent creators new ways to produce original work.
Mastercard communications and marketing chief Jill Kramer said that AI can serve as “an engine” for human creativity.
She said the technology can help marketers generate deeper insights and broaden the research and preparation that fuels new ideas, allowing them to do things they “could never do before.”
Use AI as a partner
Nosta said the solution isn’t to abandon AI but to use it as what he calls an “iterative intelligence” partner, one that helps people explore ideas rather than replacing the thinking process itself — an idea promoted by several top AI researchers, including Vivienne Ming, chief scientist at the Possibility Institute, a metascience research group.
Ming told Business Insider in March that AI is most valuable when people use it to challenge their own thinking rather than outsource it, creating what she calls “productive friction.”
“The iterative engagement is where we need to celebrate this,” Nosta said.
To him, preserving originality ultimately means preserving the human habits AI cannot replicate: curiosity, struggle, and the willingness to think differently.
“The treasure is in the tail,” Nosta said.
Read the original article on Business Insider
The post An innovation theorist says AI could squeeze out the weird ideas that change the world appeared first on Business Insider.
