Interesting take. I guess in the back of my mind I was somewhat aware that AI winter was nearly upon us. But not something I actively thought about, mainly because we still see AI ~everywhere~. Clearly being a good investor means not only following the trends, but being ahead of them. Just ask Chamath Palihapitiya. He predicated the "big-O" outcome of the Google case as a breakup of the companies. Voila, here we are, and now he already saw and likely took steps to profit from it.
Anyway, back to the topic. Gone will be the days where you can slap on an AI chatbot onto a clunky app, call it "AI enabled" and raise a ton of money. I'm noticing that VCs are getting a lot more sophisticated with their understanding of AI, and with it, comes a lack of dry powder deployed towards purely AI investments.
Especially at the angel level, this AI winter seems to have started. Is this a lead factor to the same going on in early stage, and eventually later stage investing?
Yes, extrapolation is a dangerous thing. As a stats major I understand that. And a little part of me still wonders whether the introduction of Apple Intelligence, for example, will bring about a new wave of AI innovation given in the past it is clear that huge leaps in progress are very often a lag factor of Apple doing it and showing the market that yes, there is indeed demand for a piece of tech and yes, there is a way to make it practical for consumers (think about how every mass consumption phone is not in some way, shape, or form a touch screen). But that's where the truly applied part comes in. Truly indicating one that is actually useful and pragmatic.
What to do about it? Well, here at Astra, I am trying my best to make sure that whenever we add "AI" to any of the features, it truly is useful to the consumer, not just a bell / whistle. Any time any consultant comes up with a new AI feature, my very first question is, "how is it going to benefit the user?"
Keeping the end goal is mind is key. The end goal is to have an app built on top of an ML algorithm that truly changes the way you use it. How is it actually applied and is the application really worth spending the capex AND opex? In today's day and age, AI is in an interesting time when it is seeing a lot of investor interest, but it is also actually more expensive to implement that most people probably think.
The companies who figure out how to strike that balance will be real winners imo. The rest of them risk freezing and turning into the new CPG. Still a massive TAM, but with tapered growth.