AI infrastructure faces a double-sided profit squeeze
The bearish thesis argues that the AI build-out is heading toward a sharp correction due to falling software pricing power and unrecognized, rising data center operating costs.
The argument
The guest argued that hyperscalers are facing top-line pressure as customers migrate to cheaper, highly efficient Chinese open-source models and smaller local models. Simultaneously, the true, massive costs of running these data centers are not yet fully reflected, threatening to squeeze margins from both directions.
The thesis, stress-tested
✓ What validates it
- ✓Further price cuts of 50% or more by major LLM providers
- ✓Hyperscalers reporting compressed operating margins due to power and infrastructure costs
▸ Risks discussed
- ▸Government subsidies could artificially prolong the unprofitable build-out
- ▸A sudden breakthrough in monetization could justify current CapEx levels
Hear it yourself
"something about it. And that's what he's not alone. So even the hyperscalers themselves are starting to question, the amount of token spending. They go you know, we had the surge of what they call token maxing that began when when we we started with a agentic AI really beginning of this year, with Cloud Code and those things. And, there was an explosion of of tokens, token use, and costs."
12:09 · 12:09
AFFILIATE LINK · ZORTIX MAY EARN A COMMISSION · NEVER A RECOMMENDATION TO TRADE