Amodei's AI safety call gains Altman and Musk backing

The weekend news cycle around AI doom continued as OpenAI's Sam Altman and XAI's Elon Musk joined Anthropic CEO Dario Amodei's call for independent reviewers in AI labs. But the article points out that Amodei's proposal mostly amounts to things every AI lab needs to do anyway, for commercial reasons: operational excellence, alignment, interpretability, and testing. It compares the situation to an auto CEO suggesting cars should react predictably when you turn the steering wheel.

The commentary questions Amodei's reliance on model capabilities as a danger measure, noting that if capability alone were the trigger, the panic would have started 29 years ago when Deep Blue beat Garry Kasparov. It also argues that recursive self-improvement isn't the danger either, because AI models would need to operate independently on tiny computers and learn in secret to truly threaten humanity — and frontier labs spending hundreds of billions on data centers aren't building for that.

The piece leaves open the question of what happens if an independent evaluator decides a frontier lab should shut down. Investors could be left holding the bag, and a lab might simply find a different evaluator. The real concern, the article suggests, is that hysteria may replace thoughtful debate as the AI safety story spreads from San Francisco to Washington and Brussels.