Sunday Edition: Yes True Innovation with AI is Possible—Here's the Proof

So the New York Times decided to spend Friday getting me vaguely riled up and that motivated me in a positive way to start reflecting on the fact that the innovation side of the AI story isn’t told very well.

Like honestly most of what passes for innovation news is viewed through self-serving press releases from big companies, and everyone claims everything is innovative, so what actually is?

Are there examples where we can reasonably say AI has already extended the cutting edge for humans? Where our horizon of knowledge is wider because of AI already? Because if so, that’s a really big deal.

But there’s not really been a comprehensive list of those kinds of advances in one place. Until now. I made one so at least we have a record of what looked innovative to us humans in 2025. I have a feeling this space is going to accelerate fast, and I’d like to write down some of those advances so we can check back in a few months and see what’s changed.

I don’t usually do a Sunday piece, but this felt like a nice blue sky read—sit back, reflect on the possibility of AI innovating, think a bit about what that means for us and how AI innovation changes the way we think about our own species’ role in all this stuff.

I’ll give you my cheat code there—if you look a bit more carefully at the way a lot of these innovations are getting done by AI, I don’t view them as inherently human discoverable at all, and I think that presents the interesting possibility that humans and AI have fundamentally different innovation modes that may prove complimentary if we allow that.

Sunday thinking!