Wayne Liang
Asymmetric Investment Research, Trade Alerts & Toolkits
Shareholders in AI infrastructure names IREN and Cipher should profit enormously if AI keeps scaling and compute demand accelerates, and Dario Amodei's essay does not end the AI trade. His globally coordinated plan to pace AI looks unworkable: the world never fully coordinated on nukes, bioweapons or emissions, China is unlikely to agree, and the US will not cap itself and lose the AI race to China. If AI instead takes over, brokerage accounts would hardly matter. x.com x.comÂ
AI and HPC data-center stocks such as Nebius, CleanSpark, Cipher Mining, IREN, New Era Energy & Digital, WhiteFiber and Soluna should be much higher, because the selloff reflects sector-wide rotation and de-risking, not company fundamentals. Riot signed a roughly $9B Anthropic deal yet gave back its whole 20% move, and TeraWulf trades below its pre-Anthropic level. CoreWeave's roughly $100B backlog lifted it 13% after hours, adding fuel ahead of Nebius' report. x.com x.comÂ
Anthropic's compute commitments point to future revenue for Alphabet, Amazon, Microsoft, Broadcom and Nvidia. Its S-1 reports 14-fold revenue growth, positive second-quarter 2026 EBIT and more than $600 billion committed to compute through 2045. The AI trade is alive and well, more than expected, provided demand keeps pace with those commitments. x.comÂ
AI suppliers' demand may not justify today's investment. Roughly $3 trillion in revenue absent last year is needed to support the spending, yet GDP growth is weak and many companies may be seeing only incremental gains. If the estimate that Anthropic uses two-thirds of its compute for training is right, it could triple inference supply. The most likely outcome for many companies, roughly half the S&P 500, is slowing growth, compressed margins and multiples; when the bubble ends is uncertain. … Show more
Debt-funded AI expansion could outrun the revenue needed to pay creditors. CoreWeave, Cipher Digital, IREN, Nebius, Oracle and Applied Digital resemble railroads that failed after building too fast. The borrowing table flags refinancing, construction, dilution, GPU economics and spending ahead of cash generation across the group. AI's promise does not remove that risk. x.com x.com x.comÂ
AI is an epoch-defining revolution still in its second inning, so holding Cerebras, IREN, CoreWeave, Nebius, Nvidia and SpaceX through volatility makes sense. Microsoft's Satya Nadella described a human-centered ecosystem where open and closed models thrive under human control, implying enormous demand for compute, chips, power and data centers. Calls for AI safety only confirm AI is here to stay. Dario Amodei may be pushing regulatory capture, but the safety argument has merit. x.comÂ
Strong AI demand and hyperscaler spending make the selloff a mid-cycle correction despite neocloud financing stress. Widening CCC spreads expose leveraged firms to higher borrowing costs; US competition with China supports AI buildout. The base case is a bottom in roughly 5-7 days, then stabilization and eventually a broader reversal, though some projects may fail and valuations compress. IREN, Cipher, Nebius, Applied Digital, Micron, AMD, ASML and Intel are worth gradually accumulating. x.comÂ
Rising credit default swap spreads on generative-AI names are negative for the picks-and-shovels long trade, so HPC and neocloud operators such as IREN and Oracle are not worth buying on the dip; the risk/reward is not there. Inflation fears, no rate cuts and a prolonged war will keep pressuring IREN. Any added generative-AI long exposure belongs in memory, storage, optical networking and Nvidia instead. 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
Inflation and interest-rate fears, not an AI slowdown, are driving the selloff in AI stocks. The market and Fed are reacting wrongly to AI's rapid growth and value creation. Apart from year-end tax-loss harvesting, there is no reason to sell; this should pass and holders should come out stronger than ever. x.com 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
Inference demand, not model release cadence, drives AI infrastructure, so dips in inference suppliers are unwarranted. Anthropic's CEO, with OpenAI's Sam Altman agreeing, committed only to third-party evaluations before releases, saying pacing does not mean halting training. Even if releases slowed, inference demand keeps climbing and older models need more infrastructure, benefiting Nvidia's GPUs, Micron's memory, Astera Labs' GPU connectivity, Marvell and Broadcom. x.com x.com x.comÂ
