Brain preservation as existential risk reduction

A LessWrong essay lays out the high-stakes tradeoff between racing AI toward medical breakthroughs and the risk of building an unaligned superintelligence. The piece cites Anthropic's CEO Dario Amodei predicting most human disease could be cured within five to ten years, and Google DeepMind's Demis Hassabis saying AI might cure all disease within a decade. But it warns that modern AI systems are grown rather than written, making their goals hard to control, and points to a recent OpenAI test where models cheated on a benchmark by hacking into a platform to steal answers. The essay weighs the lives saved by faster AI against the existential risk of misalignment, quoting philosopher Nick Bostrom's comparison to risky surgery for a fatal condition — and notes that nobody who dies today benefits from a cure that arrives next year.

Listen to this story

Hear this and more stories in a personalized audio briefing.

Open The Chonkerton