Ben Pouladian
Founder of BEP Research, AI Compute Economics
Open models still need GPUs, memory, networking and power, so cheaper intelligence can bring more workloads to hyperscalers' paid infrastructure. Gavin Baker notes that OpenAI and Anthropic's revenue misses parts of the AI economy. BEP's OpenRouter chart shows rising token use for selected open models; AWS, Microsoft and Google Cloud are illustrative beneficiaries, not verified hosts of those volumes. x.comÂ
Google Cloud can capture spending through existing customer cloud budgets as frontier models become interchangeable. OpenAI's GPT-6 models entered select-enterprise private preview on Google Cloud, with general availability expected in the coming weeks; Google says its catalog also offers Gemini and Claude. Pricing power moves from model labs to clouds that collect whichever model wins, and Alphabet may win more than OpenAI. x.com x.com x.com x.com x.com x.comÂ
The AI trade's biggest risk is not revenue concentration in Anthropic and OpenAI but whether scaling laws keep delivering better models that unlock valuable use cases. Steve Eisman, quoted by @firesidealpha, sees an OpenAI failure as the big risk, but that is like calling Amazon's dominance e-commerce's top risk 15 years ago. If Chinese labs or open models take share, they still need semis, and US enterprises' open models will mostly be hosted domestically by Amazon and others. x.comÂ
Microsoft's Azure has lost its OpenAI-model access edge because those models are also on AWS and are coming to Google Cloud. Azure now has to win on price and service rather than access. x.comÂ
AI infrastructure bottlenecks could give eight stocks as much as 18x upside. Nvidia supplies frontier-model GPUs; Broadcom and Marvell supply custom silicon and networking; TSMC makes advanced chips; Micron supplies sold-out HBM; Vertiv provides power and cooling; Palantir monetizes enterprise AI; Alphabet spans TPUs, research, cloud and apps. The list tracks where AI infrastructure spending is flowing. x.com x.comÂ
Anthropic's customer spending and compute buildout show AI monetizing at scale, a bullish read-through for Nvidia, Broadcom, Alphabet, Amazon, Microsoft, Nebius, CoreWeave and Hut 8. Its IPO filing reports H1 revenue up 14-fold to $16.3B, 11,900+ customers spending $100K+ annually and 1,900+ spending $1M+, alongside $9B in H1 infrastructure spending and about $430B in future compute commitments. x.comÂ
AI hyperscalers' heavy bond issuance can crowd out the corporate AI spending they depend on. After hyperscalers flooded the bond market all summer, a company rolling its Covid-era debt 5% higher has less left for an AI budget. And if every valuation and megatech management guide rests on commitments from Anthropic and OpenAI, they rest on unaudited financials never reported to the SEC. x.com x.comÂ
NVIDIA's free-model plan could defend chip demand as the two largest AI labs cut reliance on its hardware. OpenAI is building a chip with Broadcom; Anthropic mostly uses Google and Amazon chips. NVIDIA is putting together Reflection AI (in talks), Hugging Face and Groq for free models and paid chips. At 11x expected 2029 earnings versus ~19x for Microsoft, Alphabet and Amazon and 32x for Apple, its price reflects fears of slower AI spending and lost customers. NVIDIA is a 'Maximum Opportunity'. … Show more
Alphabet's shift to renting neocloud GPUs threatens significant margin compression while AI monetization remains in doubt. Analysts say Google has yet to find a way to monetize AI. Google said it began using third-party compute at the end of Q3, will ramp usage in Q4 and expects near-term margin headwinds and significantly higher capex in 2027. x.comÂ
Amazon's AWS AI rents look limited because NVIDIA backs neoclouds and Anthropic can set AWS's cut of Bedrock Claude tokens. Those tokens run on AWS-funded Trainium. Enterprise AI earns ~2 points above capital cost in the 2035 base case; shared hyperscaler advantages and new rivals limit upside, though 1-5 points is plausible. Retail ads pose a fat-tail risk despite a base case that Amazon figures them out. Market skepticism about AWS capex has toned down the bearish view. x.comÂ
Anthropic's customer growth gives AI compute suppliers firmer demand evidence than capacity plans. Its S-1 reports $16.3B in H1 2026 revenue and about 900% net retention as of September 30, the strongest evidence yet of adoption. Over $500B in commitments include a roughly $125B five-year Broadcom equipment lease plus Amazon, Google and Microsoft arrangements; Nvidia GPUs are in its mix. Long-term obligations could outlast usage revenue. x.comÂ
