The insurance angles on AI are fascinating to me
One of the more consequential disclosures in Anthropic’s S-1:
“With the scale of harm that could result from a catastrophic AI-related incident, no single insurer, and potentially the insurance industry in the aggregate, may have sufficient capacity or expertise to underwrite insurance commensurate with such exposure.”
If the whole insurance industry may not be able to underwrite frontier Al risk, we probably need something new: pooled, public-private, or built from scratch. Would love to meet anyone thinking seriously about this.