AI Spend Hits Record Highs While Use Cases Still Lag
Big Tech pours billions into models as one side sees dot-com math and the other sees supply-chain proof of real progress.
Two X accounts clash over whether current AI capex reflects sustainable gains or classic overinvestment. One flags widening hype-utility gaps and upcoming price hikes; the other points to memory shortages and agentic gains as signs the technology is still accelerating.
Why these scores — @bitts_o1 offers no revenue or usage metrics to support price-hike claims. @inNovaeight cites memory constraints without showing shipment or utilization data. Both rely on public anecdotes rather than primary financials, keeping authenticity moderate.
Nvidia added more market value in 2023-2025 than the entire GDP of Canada while most enterprise AI pilots still show single-digit ROI. That gap is the entire argument.
Side A notes that consumer features remain gimmicks and warns that firms will soon raise API prices once growth slows. Side B counters that persistent HBM shortages and early agent deployments prove demand is outrunning supply rather than collapsing.
Neither account supplies audited revenue figures or cohort retention data. The fight stays at the level of competing narratives built on public chip orders and cherry-picked demos.
Hype-utility disconnect will force higher prices once growth stalls and ROI stays flat.
- @bitts_o1✓ verified“Disconnect between hype and utility means companies will soon charge higher prices for AI features.”
Memory shortages plus agentic advances show spending is tracking real acceleration.
- @inNovaeight“Persistent memory shortages and agentic advances show AI is accelerating despite short-term lags.”
Read it straight — Pull Microsoft and Google 10-Q filings and compare AI-linked revenue lines against capex totals for the same quarters.