richard craig

Numerai demonstrates that sustainable returns come from hedging thousands of risk factors to isolate pure alpha, using leverage to scale a low-vol core; concentrated AI funds like Situational Awareness take factor risk (beta) not alpha, leading to catastrophic drawdowns when regimes shift.

Directional funds conflate factor risk (market, momentum, tech) with alpha; sustainable returns come from neutralizing thousands of risk factors, using leverage only on residual alpha, and targeting 9% annual alpha compounded over decades rather than 400% in months with catastrophic drawdown risk.

Investors can rationally allocate to high-volatility funds like Situational Awareness if sized appropriately within a diversified portfolio, as losses in one fund may be offset by gains elsewhere.

The stock market itself functions as an artificial intelligence that efficiently prices AI narratives; investors who simply buy AI stocks because 'AI will be big' are competing against algorithmic market-makers and should not take the thesis lightly.

Since AI progress is already reflected in market prices and the market itself functions as an AI, simply betting on AI stocks adds factor risk not alpha.

AI model economics are shifting from cost-per-token to cost-per-task, with token efficiency becoming a key differentiator as closed-source models undercut open-source on price.

Benchmark results for frontier models like OpenAI's o3 depend heavily on harness settings (e.g., memory across turns), with correct configuration tripling scores and reducing token usage 6x.