Author
@scaling01
grade Cx.com/scaling01 ↗
@scaling01 is a pseudonymous AI/tech critic with large engagement, and some posts touch AI-company valuations or fundraising, but the feed is mainly AI benchmarks, product commenta
trader score
-1.58
hit rate
40%
mean α
-0.78%
signals 14d
57
Grade = how their written analysis reads (A best). Trader score = how their last-20 timestamped calls performed vs SPY.
Their picks, scored
Across their last 20 scored bets: 40% hit rate, -0.78% mean alpha, trader score -1.58. Their last-14d mentions, direction-adjusted, have moved -1.5% since posting (mean over 37 mentions with price data).
Recent signals15receipts included
| date (PT) | ticker | author | sent | what they said | since then | receipt |
|---|---|---|---|---|---|---|
| 2026-10-16 | · | @scaling01 | · | Announces a research project on looping models. | · | tweet ↗ |
| 2026-10-15 | BTC | @scaling01 | +0.40 | Expresses longstanding enthusiasm for Tesla and regret over missed early investments in TSLA, AMD and BTC. | · | tweet ↗ |
| 2026-10-15 | AMD | @scaling01 | +0.40 | Expresses longstanding enthusiasm for Tesla and regret over missed early investments in TSLA, AMD and BTC. | -8.1% | tweet ↗ |
| 2026-10-15 | TSLA | @scaling01 | +0.65 | Expresses longstanding enthusiasm for Tesla and regret over missed early investments in TSLA, AMD and BTC. | +2.7% | tweet ↗ |
| 2026-10-15 | · | @scaling01 | · | Claims Mistral has reached the capability level of frontier AI models from April 2025. | · | tweet ↗ |
| 2026-10-15 | AAPL | @scaling01 | +0.00 | Disputes a claim that Apple-OpenAI files concern model or training secrets leaked to China. | +1.7% | tweet ↗ |
| 2026-10-15 | GOOGL | @scaling01 | -0.35 | Says Gemini 4 Argon benchmarks well broadly but performs poorly on reliability and hallucinations. | -8.3% | tweet ↗ |
| 2026-10-15 | MSFT | @scaling01 | +0.40 | Sarcastically rejects claims China has reached the frontier, arguing it remains behind Microsoft. | -2.8% | tweet ↗ |
| 2026-10-15 | MSFT | @scaling01 | +0.25 | Uses MAI-Base-2's 16,000 GB200 GPUs and 38-day run to argue frontier pretraining can be fast. | -2.8% | tweet ↗ |
| 2026-10-15 | NVDA | @scaling01 | +0.20 | Uses MAI-Base-2's 16,000 GB200 GPUs and 38-day run to argue frontier pretraining can be fast. | -2.9% | tweet ↗ |
| 2026-10-15 | MSFT | @scaling01 | +0.35 | Claims earlier high rankings for Meta and Microsoft were vindicated by their compute advantage over China. | -2.8% | tweet ↗ |
| 2026-10-15 | META | @scaling01 | +0.35 | Claims earlier high rankings for Meta and Microsoft were vindicated by their compute advantage over China. | -12.8% | tweet ↗ |
| 2026-10-15 | META | @scaling01 | +0.00 | Compares Meta's suspected model architecture with disclosed Muse Spark parameter and layer counts. | -12.8% | tweet ↗ |
| 2026-10-15 | · | @scaling01 | · | Attributes an ARC-AGI-3 capability jump mainly to game reinforcement learning rather than looping. | · | tweet ↗ |
| 2026-10-15 | · | @scaling01 | · | Predicts sufficiently resourced AI systems will solve difficult interpretability and connectomics problems. | · | tweet ↗ |
Grade is our human read-worthiness rating; trader score is a rolling 20-bet hit-rate/alpha composite — different things, often disagreeing. “Since then” is direction-unaware in the table; the summary line above adjusts for which way they leaned.