Dario Amodei directly contradicts the belief that his preferred regulatory path has failed, asserting that recent events have not derailed it.

Scaling laws are empirical relationships observed for particular architectures, objectives, and datasets, not laws of physics, meaning changing any one of those factors produces a different scaling curve.

Brian Greene puts a hard ceiling on the assumption that AI scaling is automatic

The AI buildout is primarily a debt cycle, not an equity cycle.

AI will lead to more software engineers in the world overall, even though fewer will be needed to do existing work.

The current market irrationality is concentrated in earnings expectations, analysts are systematically too optimistic about earnings, rather than in valuations.

The AI model available today will be the worst model ever made, as every future model will be better, implying continuous, unending improvement.

A relatively small reduction in R0, from just above 1 to just below 1, is the difference between a pandemic and no pandemic.