{"version":"market-claims.v1","expert":{"slug":"andrewdfeldman","displayName":"Andrew Feldman","kind":"PERSON","status":"UNCLAIMED","avatarUrl":"/api/avatars/923c7a66-3","url":"/trace/experts/andrewdfeldman","claimed":false},"stats":{"sources":2,"marketClaims":2,"explicitCalls":0,"directionalTheses":2,"ambiguousMentions":0,"openClaims":2,"median30dReturn":null,"lastUpdatedLabel":"recently"},"claims":[{"id":"mcl_203d406ff90d14c346","url":"/trace/claims/mcl_203d406ff90d14c346","status":"PUBLISHED","claimType":"DIRECTIONAL_THESIS","direction":"SHORT","claimText":"A moat that turns out to be optional drains the premium out of the most expensive name in AI: frontier models now train with zero CUDA, and the pricing power that justifies the valuation erodes as buyers gain real alternatives.","confidence":0.72,"expert":{"slug":"andrewdfeldman","displayName":"Andrew Feldman","kind":"PERSON","status":"UNCLAIMED","avatarUrl":"/api/avatars/923c7a66-3","url":"/trace/experts/andrewdfeldman","claimed":false},"primaryEntity":{"kind":"ASSET","assetClass":"EQUITY","symbol":"NVDA","name":"NVDA","exchange":null,"confidence":0.78},"evidence":{"evidenceUrl":"https://www.youtube.com/watch?v=2zWrLqo7rLQ&t=9193s","sourceUrl":"https://youtube.com/watch?v=2zWrLqo7rLQ","timecode":"153:13","startMs":9193000,"endMs":null,"excerpt":"Yeah. And I think trying to share what that means, what an exponential demand means, and that we're still so early and yet the demand for AI compute is overwhelming. I think sharing that was interesting and I think helpful in educating the financial community. The other thing is that there are lots of ways to do this. That the GPU isn't the only way. You've got a TPU, you've got Trainium, you've got us."},"performance":{"entryPrice":235.04,"currentPrice":null,"returnPct":null,"relativePct":null,"method":"PASTE_AUTHOR_PRICE"}},{"id":"mcl_a493d8e80082ed0273","url":"/trace/claims/mcl_a493d8e80082ed0273","status":"PUBLISHED","claimType":"DIRECTIONAL_THESIS","direction":"LONG","claimText":"Cerebras' dinner-plate die with memory next to compute runs inference 15-18x faster than GPUs, a new silicon winner","confidence":0.72,"expert":{"slug":"andrewdfeldman","displayName":"Andrew Feldman","kind":"PERSON","status":"UNCLAIMED","avatarUrl":"/api/avatars/923c7a66-3","url":"/trace/experts/andrewdfeldman","claimed":false},"primaryEntity":{"kind":"ASSET","assetClass":"EQUITY","symbol":"CBRS","name":"CBRS","exchange":null,"confidence":0.78},"evidence":{"evidenceUrl":"https://www.youtube.com/watch?v=jLICvWE7w2Q&t=1320s","sourceUrl":"https://youtube.com/watch?v=jLICvWE7w2Q","timecode":"22:00","startMs":1320000,"endMs":null,"excerpt":"we made two bets. Um, the first was dedicated silicon would be the answer, and the second was it couldn't look like a GPU. And our view as computer architects is if you want to be 20 times better than somebody Right, your architecture can't look like them"},"performance":{"entryPrice":241.27,"currentPrice":null,"returnPct":null,"relativePct":null,"method":"PASTE_AUTHOR_PRICE"}}]}