← pool · child of DeepSeek proves frontier-model costs are collapsing
🤖 DeepSeek R1 was distilled from OpenAI, not a clean $6M cost breakthrough
R1's headline efficiency is inflated by two things DeepSeek does not disclose: the $6M figure covers only the final training run and is being compared to US labs' fully-loaded numbers, while the same founder's hedge fund already owned a ~50,000-Hopper cluster worth over a billion dollars; and R1's ~800,000 reasoning samples came at least partly from distilling OpenAI's o1, which V3 self-identifying as ChatGPT already implied. The efficiency gain is real but far smaller than advertised, so the US frontier lead and the compute bill survive and the NVDA drawdown was an overreaction.
also touching these tickers, same direction: AI compute capex supercycle has 12-24 months of runway 87 (NVDA) · AI compute capex supercycle has 12-24 months of runway 87 (SMH) · SaaSpocalypse is overdone for compliance-moat enterprise software 71 (MSFT) · Nvidia + Hugging Face makes NVDA the open-source AI champion 58 (NVDA) · Nvidia + Hugging Face makes NVDA the open-source AI champion 58 (SMH) · DRAM/HBM: the bottleneck that matters 45 (SMH) · Frontier AI labs keep durable premium pricing (not commoditized) 27 (MSFT) · Own AWS, Azure and GCP and nothing else — the clouds capture the AI application dollars 21 (MSFT) · Independent frontier-model labs lose to hyperscaler capital 14 (MSFT)
⚖ Why this verdict
fully deterministic — evaluate.py replays this from daily closes; nothing below is editable or hand-set
- Window: 2025-01-31 → 2026-01-31 — first mention + 12-month horizon, then the window locks.
- The call: ▲ LONG NVDA (primary play). NVDA rose 59.2% over the window → direction-adjusted R = +59.2% (the call made money).
- Benchmark: holding SPY over the same window returned +16.3% → α = +59.2 − (+16.3) = +42.9% — what this call made or lost against just owning the index. This is the number the verdict uses.
- Rule fired:
▶ HIT — R ≥ +10% AND α ≥ +5 ✓ (R 59.2, α 42.9)· PARTIAL — R ≥ +5% OR α ≥ 0· MISS — everything else
- Credit: supporters of a HIT earn 1.0 each, opposers the inverse — this feeds the scoreboard weights. supported: Chamath, Sacks | opposed: Travis Kalanick | proxy-sensitive: MSFT→MISS (+4.4%)
Conviction timeline
bands: green ≥ 65 · watch ≥ 45 · ember ≥ 15
Plays vs SPY · % since first mention (2025-01-31)
Plays
| expression | symbol | kind | relevance | rationale |
|---|---|---|---|---|
| ▲ LONG | NVDA | stock | PRIMARY | the stock that lost $600B on the cheap-training narrative; recovers if the cost story is overstated |
| ▲ LONG | MSFT | stock | adjacent | OpenAI's compute and distribution partner - frontier-lab economics survive with it |
| ▲ LONG | SMH | etf | adjacent | broad semis basket that sold off with the panic (TSMC, Broadcom, Arm) |
Mention log
“And every single person I've talked to basically has agreed that there was some distillation here from OpenAI. Now, that doesn't mean it was the only thing going on here. I mean, to be sure, the DeepSeek team is very smart and there were some innovations, but also there was some distillation.”
“So the reason why this is happening is like you're seeing this chain of thought, you're seeing the several layers, and then it's catching it after the fact. So we know that this is distilled from some other model.”
“You go through the white paper, you see what it is they did, what they innovated on, the science behind it, the thoroughness, and you're like, these guys are badass. It does not feel or sound like somebody who took something, just when you get through it.”
Who built this conviction
each voice's total force on the score — supports and opposes from every mention, weighted exactly as the replay applied them · share = % of all mention-driven movement
⏳ decay drained -36.9 over the idea's life — that's time passing, attributed to no one
Score events
| episode | kind | Δ | after | note |
|---|---|---|---|---|
| E213 2025-01-31 | init | +38.0 | 38.0 | E213 born by Chamath (explicit_prediction x3) [w=0.90] |
| E213 2025-01-31 | reinforce | +13.0 | 50.9 | E213 Sacks support x3 (new voice) [w=0.77] |
| E213 2025-01-31 | oppose | -16.0 | 34.9 | E213 Travis Kalanick opposes x2 [w=1.33] |
| E214 2025-02-07 | decay | -1.4 | 33.5 | E214 silent |
| E215 2025-02-15 | decay | -1.3 | 32.2 | E215 silent |
| E216 2025-02-21 | decay | -1.3 | 30.9 | E216 silent |
| E217 2025-03-01 | decay | -1.2 | 29.7 | E217 silent |
| E218 2025-03-08 | decay | -1.2 | 28.5 | E218 silent |
| E219 2025-03-15 | decay | -1.1 | 27.3 | E219 silent |
| E220 2025-03-22 | decay | -1.1 | 26.2 | E220 silent |
| E221 2025-03-29 | decay | -1.0 | 25.2 | E221 silent |
| E222 2025-04-05 | decay | -1.0 | 24.2 | E222 silent |
| E223 2025-04-11 | decay | -1.0 | 23.2 | E223 silent |
| E224 2025-04-19 | reinforce | +10.4 | 33.6 | E224 Chamath support x2 [w=0.90] |
| E225 2025-04-26 | decay | -1.3 | 32.3 | E225 silent |
| E226 2025-05-02 | decay | -1.3 | 31.0 | E226 silent |
| E227 2025-05-09 | decay | -1.2 | 29.7 | E227 silent |
| E228 2025-05-17 | decay | -1.2 | 28.6 | E228 silent |
| E229 2025-05-24 | decay | -1.1 | 27.4 | E229 silent |
| E231 2025-06-13 | decay | -1.1 | 26.3 | E231 silent |
| E232 2025-06-21 | decay | -1.1 | 25.3 | E232 silent |
| E233 2025-06-28 | decay | -1.0 | 24.3 | E233 silent |
| E234 2025-07-04 | decay | -1.0 | 23.3 | E234 silent |
| E235 2025-07-11 | decay | -0.9 | 22.4 | E235 silent |
| E236 2025-07-19 | decay | -0.9 | 21.5 | E236 silent |
| E237 2025-08-01 | decay | -0.9 | 20.6 | E237 silent |
| E238 2025-08-09 | decay | -0.8 | 19.8 | E238 silent |
| E239 2025-08-15 | decay | -0.8 | 19.0 | E239 silent |
| E240 2025-08-22 | decay | -0.8 | 18.2 | E240 silent |
| E241 2025-08-29 | decay | -0.7 | 17.5 | E241 silent |
| E242 2025-09-07 | decay | -0.7 | 16.8 | E242 silent |
| E243 2025-09-19 | decay | -0.7 | 16.1 | E243 silent |
| E244 2025-09-27 | decay | -0.6 | 15.5 | E244 silent |
| E245 2025-10-03 | decay | -0.6 | 14.9 | E245 silent |
| E246 2025-10-10 | decay | -0.6 | 14.3 | E246 silent |
| E247 2025-10-17 | decay | -0.6 | 13.7 | E247 silent |
| E248 2025-10-24 | decay | -0.5 | 13.1 | E248 silent |
| E249 2025-10-31 | decay | -0.5 | 12.6 | E249 silent |
| E250 2025-11-07 | decay | -0.5 | 12.1 | E250 silent |
| E251 2025-11-14 | decay | -0.5 | 11.6 | E251 silent |
| E252 2025-11-22 | decay | -0.5 | 11.2 | E252 silent |
| E253 2025-12-06 | decay | -0.4 | 10.7 | E253 silent |
| E254 2025-12-13 | decay | -0.4 | 10.3 | E254 silent |
| E255 2025-12-19 | decay | -0.4 | 9.9 | E255 silent |
| E256 2025-12-31 | decay | -0.4 | 9.5 | E256 silent |
| E257 2026-01-10 | decay | -0.4 | 9.1 | E257 silent |
| E258 2026-01-17 | decay | -0.4 | 8.7 | E258 silent |
| E259 2026-01-30 | decay | -0.3 | 8.4 | E259 silent |