Innovation through the wrong end of the telescope

New study on AI use in 41M scientific studies: Scientists using AI publish 3x more papers, get 4.8x more citations, become team leaders 1.4 years earlier. But the breadth of subjects studied shrank 5%. Collaboration dropped 22%.

Everyone is sprinting toward the same finish line.

Now your innovation team has the same problem.

This study found the AI paradox nobody's talking about:

individual impact explodes while collective exploration contracts.

Think of it like Moneyball.

Every team adopted the same analytics.

Every GM started drafting the same undervalued players.

The edge disappeared the moment everyone found it.

That's what's happening inside most companies right now.

Teams are shipping features faster. Building dashboards quicker. Answering customer questions at scale. But when I map organizational capability across companies, everyone's optimizing the exact same things: backoffice functions, customer journey touchpoints, AI-powered features.

Everyone’s fishing in the same data-rich ponds because that's where AI performs best.

Meanwhile, the hard innovation problems—the ones requiring exploration over optimization—are getting systematically deprioritized. The pipeline's narrowing precisely when you think it's expanding.

The researchers say, "AI automates established fields rather than explores new ones."

In business, that's commoditization.

The fix isn't using less AI. It's deploying it strategically, not opportunistically. That means protecting exploration capacity even as you automate execution. Building org structures that resist convergence. Measuring breadth of experimentation, not just the velocity you’re shipping.

I'd bet companies winning in 2026 won't be the ones who adopted AI fastest.

They'll be the ones who figured out where NOT to use it.

—

Source: Artificial Intelligence Tools Expand Scientists' Impact but Contract Science's Focus, Hao et al, Nov 2025