Gen AI Won't Prevent The Economy from Rolling Over
Too many investors believe that “AI” will step into the breach and pick up the slack for a sagging economy. That belief is wrong.
First, broadly-defined “AI” has been with us for decades. The new wrinkle is the cross domain large language and multi-modal models that companies such as OpenAI, Google, Microsoft, Amazon and Meta Platforms have made significant investments in.
The problem is that Gen AI models are not end-user products. API services are not end-user products. I’ve yet to see Gen AI successfully productized.
Microsoft’s Copilot Pro product has done anything resembling lighting the world on fire. That’s one big Gen AI barometer that’s pointing toward failure.
The first Gen AI product set will likely be super-powered AI agents (think Google Assistant on steroids, Siri on steroids or a version of ChatGPT), yet those agents won’t account for 1% of GDP steady state.
Gen AI will take years to evolve from a product standpoint.
I believe the pace of Gen AI investment is about to pull back.
VCs are not going to write large Gen AI checks if they believe that Gen AI valuations are about to get hit and/or if Gen AI portfolio companies are suffering (they are unless the name on the door is OpenAI, Anthropic or Mistral).
MSFT, GOOG, META, AMZN, OpenAI and others will pull back on the pace of Gen AI investment if they see the number of corporate experiments dwindle (that is happening) and/or if the economy pulls back and it becomes an earnings protection game.