“We have to do more with less” has become an inescapable mantra, and goodness, are you trying.  You’ve slashed projects and budgets, “right-sized” teams, and tried any technology that promised efficiency and a free trial.  Now, all that’s left is to replace the people you still have with AI creativity tools.  Welcome to the era of the AI Innovation Team.

It sounds like a great idea.  Now, everyone can be an innovator with access to an LLM.  Heck, even innovation firms are “outsourcing” their traditional work to AI, promising the same radical results with less time and for far less money.

It sounds almost too good to be true.

Because it is too good to be true.

AI is eliminating the very brain processes that produce breakthrough innovations.

This isn’t hyperbole, and it’s not just one study.

MIT researchers split 54 people into three groups (ChatGPT users, search engine users, and no online/AI tools using ChatGPT) and asked them to write a series of essays.  Using EEG brain monitoring, they found that the brain connectivity in networks crucial for creativity and analogous thinking dropped by 55%.

Even worse? When people stopped using AI, their brains stayed stuck in this diminished state.

University of Arkansas researchers tested AI against 3,562 humans on a series of four challenges involving finding new uses for everyday objects, like a brick or paperclip.   While AI scored slightly higher on standard tests, when researchers introduced a new context, constraint, or modification to the object, AI’s performance “collapsed.” Humans stayed strong.

Why? AI relies on pattern matching and is unable to transfer its “creativity” to unexpected scenarios. Humans use analogical reasoning so are able to flex quickly and adapt.

University of Strasbourg researchers analyzed 15,000 studies of COVID-19 infections and found that teams that relied heavily on AI experts produced research that got fewer citations and less media attention. However, papers that drew from diverse knowledge sources across multiple fields became widely cited and influential.

The lesson? Breakthroughs require cross-domain thinking, which is precisely what diverse human teams provide, and, according to the MIT study, AI is unable to produce.

How to optimize for efficiency AND impact (and beat your competition)

While this seems like bad news if you’ve already cut your innovation team, the silver lining is that your competition is probably making the same mistake.

Now that you know better, you can do better, and that creates a massive opportunity.

Use AI for what it does well:

• Data analysis and synthesis
• Rapid testing and iteration to refine an advanced prototype
• Process optimization

Use humans for what we do well:

• Make meaningful connections across unrelated domains
• Recognize when discoveries from one field apply to another
• Generate the “aha moments” that redefine industries

Three Questions to Ask This Week

1. Where did your most recent breakthroughs come from? How many came from connecting insights across different domains? If most of your innovations require analogical leaps, cutting creative teams could kill your pipeline.
2. How are teams currently using AI tools? Are they using AI for data synthesis and rapid iteration? Good. Are they replacing human ideation entirely? Problem.
3. How can you see it to believe it? Run a simple experiment: Give two teams an hour to solve a breakthrough challenge. Have one solve it with AI assistance and one without.  Which solution is more surprising and potentially breakthrough?

The Hidden Competitive Advantage

As AI commoditizes pattern recognition, human analogical thinking and creativity become a competitive advantage.

The companies that figure out the right balance will eat everyone else’s lunch.