Explain how the market is backed by real earnings if end consumer doesn't benefit nearly enough from AI?

So many people justify the market right now by saying: look, market is backed by earnings. Ok, those earnings are real, but they are due to supply chain investments in AI, not from end-consumer real AI profits.

Everyone builds data centers and trains LLMs with the expectation that jobs will be replaced and productivity will increase a lot, but this didn't happen much yet, at least not compated to the investments made.

If end users don't benefit that much from AI as expected, all those data centers and everything auxilary will crash like crazy.

It's like you make 500 pies expecting people will buy them. You buy floor, apples etc, from local market - they will have an increase in earnings. What happens if those pies are not bought by people? Next time you don't need so many apples and floor, market earnings wil decrease next day.

Unless end-user shows real productivity gains with AI, all these investements and capex can mean shit, it will be temporary market surge and a big crash will come next.

Is there any proof that all this capex investment will pay off to the end user?

LE: Seems people didn't understand my take. Is that enough economic value produced by AI to justify hundreds of billions of CAPEX investments? Besides marginal increase in software and few other domains, is there any other proof to justify the expected growth in AI infrastracture?

LE2:
In a study in 2026 on 6000 senior executive, more than 80% reported AI had no impact so far on either productivity or employment for next 3 years. Only roughly:

**+1.4% productivity**
**+0.8% output**
**−0.7% employment**

LE3:
Looking for non-biased answers. If you're heavily invested in the market it's hard to not be biased. I'm personally invested, but I still see the risks