≈

The passage around this formula

This piece assembles the documented cost structure of alignment and safety work at the scale frontier labs now operate, decomposed into three cost centres that map onto three distinct kinds of scarcity: $C_{\mathrm{align}} \approx C_{\mathrm{human}} + C_{\mathrm{RL}} + C_{\mathrm{redteam}}$. $C_{\mathrm{human}}$ is the cost of collecting human feedback — demonstrations and preference comparisons, priced per label and bounded by how many people can be recruited and trained. $C_{\mathrm{RL}}$ is the compute spent training a policy against a reward model or a verifier, priced in GPU-hours and bounded by the same hardware capability development competes for. $C_{\mathrm{redteam}}$ is the cost of adversarially testing the result, priced in expert-hours and bounded by…

Read this part in the article →