The Valley of AI Innovation

Why This Feels Like 2006

Happy Sunday and welcome to Investing in AI. This week I want to talk about how things have slowed down in AI, and why, to me, it feels like 2006.

Tomasz Tunguz wrote recently about the slowdown in software markets. It’s real, and I’ve seen several other people writing about it. In one good example, Molly White recently asked if AI is worth it. It feels like a wave of “AI is overhyped” is coming.

With the exception of foundation models and a few hot AI companies, things seem to be slowing down, both on adoption and financing. It’s easy to think, at a time like this, that we are entering some valley of AI that we may not get out of for some time. But to me, this feels a lot like 2006.

I remember talking to a friend in 2006 saying it felt like there was nothing innovative coming. It was a time when there didn’t seem to be any big breakthroughs, and companies were raising on stories like “we use AJAX,” or “we are a platform to add tag clouds to websites.” The platforms that were popping up, things like Reddit and Facebook, felt like toys. None of it felt like real innovation.

But just a few years later I remember pointing out to that same friend that the reason it felt slow was, we moved from a technical innovation period through a user behavior change period. Adoption was slow and uncertain. No one knew what to do with these things, and the tools hadn’t hit any kind of mass adoption. Yet bubbling under the surface was something big.

I don’t know what is next. Maybe it’s agents, maybe it’s edge AI. I’m doubtful it’s AGI. But something is brewing and in 2 years I think we will what the right theses were for investing in this phase of AI, and what types of companies are starting to win. In some ways it feels quiet, but it’s a very exciting time.

Thanks for reading.