The $7.5 trillion argument says the AI boom isn’t anywhere near the historical spending limit yet.
Fine.
But I think the more interesting problem is happening underneath that number.
AI is getting cheaper incredibly fast while the cost of building the AI machine is still exploding.
That means the bull case now needs usage to go absolutely vertical.
If token prices keep falling another 50%, companies need a hell of a lot more AI consumption just to make up for the lost pricing.
So I’m not ready to call the AI top.
But I also wouldn’t use the $7.5 trillion number to declare everything is fine.
The real test isn’t how much they can spend. It’s whether AI revenue can grow faster than AI prices are falling.
That is where this gets interesting.
⚠️AI pricing is collapsing, raising questions about AI returns:
The LLM Token Expenditure Index, which tracks the market price companies pay for AI model output, has fallen to just $0.97, its lowest level since the index was created late last year and more than -50% below its… pic.twitter.com/fZm4KucuwG
— Global Markets Investor (@GlobalMktObserv) September 2, 2026
🚨 WARNING: BIG STORM INCOMING!
The stock market is at all-time highs.
But almost nobody is paying attention to the real problem.
The Fed is about to hike rates next month.
Here's how that turns into a systemic crash:
The S&P 500 is no longer a diversified bet on the U.S.… https://t.co/y3DfnFRAIO pic.twitter.com/WxdYChooV2
— MARMOT (@Web3Marmot) September 1, 2026
Barron’s says historically Capex cycles tend to peak at 25% of total U.S. GDP
The railroad spending peaked at $2.5B out of $10B GDP
The Dot com spending peaked at ~ $1.5T in $6T economy at the time
Same logic means danger zone for Hyperscalers is ~ $7.5T
So Party back on🤪 pic.twitter.com/rfzZ6d2p2p
— Q-Cap (@qcapital2020) September 1, 2026