What's this about?
Matt Yglesias argued it is to the credit of Sam Altman and Dario Amodei that they told the public what leading researchers in the field actually believe, and that the greater concern should be politically savvy reassurance from executives who say nothing alarming. Ethan Mollick added the sharper formulation: they talked about AI replacing all labour and posing existential risk because, as far as he could tell, they believed it long before their companies were valuable — and the fact that they stopped talking about risk is the PR part, not the earlier warnings.
What exactly is the cynical reading being rejected?
The common accusation is that existential-risk talk was always marketing: claiming your product might end the world is an extraordinary capability claim, and it attracts regulation that entrenches incumbents. The counterargument here is chronological. These people were saying it when their companies were small and the claim brought them ridicule rather than investment. If it were a strategy, it was an expensive one — and the strategic behaviour is visible in the *reversal*, as risk talk quietly disappeared once it became commercially inconvenient.
Why is this worth noticing?
- "The fact that they stopped is the PR part" inverts the usual accusation, and it's testable against the record in a way most claims about motive aren't.
- It lands the week OpenAI paused frontier training over cyber capability — an action that fits the earlier framing rather than the reassuring one.
- The point about politically savvy reassurance is the one with policy consequences. Executives who say nothing alarming attract no scrutiny, which is a poor selection mechanism for public attention.