Thread
compute-scarcity
8 reads · chronological order
We're Using So Much AI That Computing Firepower Is Running Out
The compute scarcity thesis just went mainstream: WSJ reports Anthropic's 98.95% uptime as enterprise clients defect to OpenAI, Blackwell GPUs up 48% in two months, and OpenAI killed Sora to free tokens for coding. The buried signal isn't the shortage itself; it's that Retool's CEO switching providers over reliability — not capability — previews what happens when inference demand compounds faster than infrastructure can respond. The company that solves five-nines for AI inference will own enterprise, regardless of whose model benchmarks best.
We're Using So Much AI That Computing Firepower Is Running Out
Retool's CEO switched from Anthropic to OpenAI this quarter, and the reason wasn't a benchmark: it was 98.95% uptime versus the alternative. Enterprise AI competition has shifted from capability to reliability, the same transition cloud infrastructure went through in 2010. The Anthropic paper this week shows the same pattern one layer up: automated alignment research can generate at $22/hour, but generation without stable evaluation infrastructure is just faster reward-hacking. Davies' vigilance decrement argument lands it at the human layer: even if the infrastructure holds, the person reviewing outputs degrades before the system does. Whoever solves five-nines for the full stack, model plus evaluation plus human judgment, owns enterprise regardless of whose Elo score leads.
AI is confronting a supply-chain crunch
Hyperscaler capex grew 190% from 2024 to 2026; their hardware suppliers grew 45%. That gap is why every throttling notice, plan change, and Sora shutdown traces back to the same constraint. The less-discussed dimension: agentic systems need 1 CPU per GPU versus 1:12 for chatbots, which is why Intel has doubled in six months and why every agent platform deck needs a CPU supply slide.
BG2Pod with Brad Gerstner and Bill Gurley — Jensen Huang | Does AI Revenue Justify Capex, '27 AI Bottlenecks, Singularity, Open vs Closed & more
Jensen argues the moat is growing because a $50-100B order can't be benchmarked, so buyers underwrite it on fungibility, durability and rentability. Nvidia's content per gigawatt has climbed from about $18B to 40-plus B, and roughly a third of the business is startups renting compute. That is the lender's checklist, and it explains why Hoffman's guarantee book exists at all.
BG2Pod with Brad Gerstner and Bill Gurley — Jensen Huang | Does AI Revenue Justify Capex, '27 AI Bottlenecks, Singularity, Open vs Closed & more
Jensen's claim that the moat is growing rests on a buyer's logic: a $50-100B order can't be benchmarked, so it gets underwritten on fungibility, durability and rentability. Nvidia's content per gigawatt has gone from about $18B to 40-plus B, and about a third of the business is startups renting compute, so payback does not hinge on two labs reaching trillion-dollar revenue. SemiAnalysis expects 43 GW of new AI compute next year, against a Street view of 35-40 GW of land, power and shell that may come online. That is the lender's checklist, and it explains why a guarantee book exists at all. What it adds to the week is the demand side of the financing: the AI-bubble question now sits with the non-lab two-thirds and how fast power gets energized.
Want to Short the AI Boom? She Wants to Give You a Way to Bet on 'Compute' Prices
CME plans cash-settled monthly contracts on H100 and B200 rental, out to three years, based on Silicon Data indexes built from 40+ sources. The CFTC has already flagged manipulation by dominant players, price variance across providers and regions, and missing fungibility, and its review pushed the October 5 launch back by 45 days. A traded forward curve would let lenders and shorts price the scarcity the chips and Empires of AI reads describe, but the settlement number belongs to whoever owns the index.
Empires of AI: The Rise and Fall of a New Global Order
The Empires of AI piece argues that scaling laws tie the frontier to energy, chips and capital, so capital decides who can keep racing. The US paid about $881B in interest against $874B on defense in 2024, while the biggest tech firms committed more than $600B of 2026 capex. Add the June Commerce order that took Anthropic's top two models dark worldwide, and access to the output of those scarce tools looks like a revocable state concession.
Anthropic Flags 'Existential Risks' in Long-Awaited IPO Filing
Ludlow reads the filing's risk language as saying suppliers won't give Anthropic all the compute it asks for, so the binding limit is capacity, not cash. The commitment figures don't reconcile: $250B he calls absolutely on the hook, $417B in minimum obligations running to 2046 per Shepard. With seven founders holding 50.1% of votes, IPO buyers are underwriting contract execution, which makes those terms the main diligence item.