If the AI bubble does pop, what happens next?

You are relegated having a booth at AI World 2027 in Santa Monica. You will have to dress up in branded polo shirts and shout down passers-by with claims of your AI wares. You will have custom backdrops printed. You will break even on the marketing expenditure, but just barely.

We all knew the time would come, but take a second and appreciate how things are right now. AI is currently like the early days of the internet, before ads ruined everything.

So here are a list of companies that are doing something with AI that are funded by YC.

https://news.ycombinator.com/item?id=44981253

How many of them do you think will be around in a year or have a successful exit?

As far as you idea of adding ads to a gen AI products, what customer are you targeting and why should any company use your potential product instead of just doing something in house?

What is your “unfair advantage”?

> what happens next

If we really have hit the Trough of Disillusionment then the next thing us the Slope of Enlightenment [1].

Thats when people figure out how to actually make it work.

In terms of opportunity and £££ what matters is the area under the curve.

[1] https://en.wikipedia.org/wiki/Gartner_hype_cycle

AI is not going to go away completely, but we might not see things like "Apple Intelligence" being advertised so breathlessly, not to mention Google's new phone that wants to talk with you about your environment.

The .com bubble of the 1990's would be a good model.

I think a useful analogy is to think of AI as equivalent to the Steam Engine. Instead of making muscle power obsolete, it provides cognition on an industrial scale.

The hype cycle is real, and there will be, as there was with railroad construction in the 1800s, overshoot of production. Unlike with railroads however, demand will meet supply, and just eat up the compute that's available, voraciously.

The profit margins will likely be lower for the suppliers, and higher for those using AI to build things, in the long run, if they satisfy an actual market need.

Then we might be able to pick a few cheap stocks or hide in a cabin watching nuclear blasts coming down.

Unlike the internet and mobile internet, the spread of AI doesn’t require additional hardware devices, so its conversion into productivity is happening much faster. I believe that even if the AI bubble bursts, it won’t be as destructive as the internet crash of the 2000s. Still, it’s clear that the market is showing signs of overheating. As a practitioner, I think the best strategy is to accumulate as much capital as possible during the upcycle while enforcing strict spending discipline.