We can't find the internet
Attempting to reconnect
Something went wrong!
Hang in there while we get back on track
Comment by Connor Leahy
Conjecture CEO; AI safety researcher
The primary prerequisite to even considering starting to work on a safe ASI plan is to have a global ASI ban and powerful enforcement already in place. Unsafe ASI is vastly easier to build than controlled ASI, and is on the same tech path. [...] The point of an ASI ban/pause is to create the time to reduce the gap between unsafe and controlled ASI, until it's sufficiently narrow that competent people can walk across without falling through.
AI Verified
source
(2026)
1mo ago
Policy proposals and claims
Verification History
AI Verified
Quote attributed to Connor Leahy (Conjecture CEO, AI safety researcher), from his April 2026 post "You can only build safe ASI if ASI is globally banned" (Alignment Forum / his Substack). The source_url (alignmentforum.org) returned HTTP 403 to direct WebFetch, but a web search confirmed the post and its core claim verbatim: "The primary prerequisite to even considering starting to work on a safe ASI plan is to have a global ASI ban and powerful enforcement already in place," along with his argument that unsafe ASI is easier to build than controlled ASI and that a ban/pause buys time to narrow that gap. Attribution correct, year 2026, relevant. Vote "for" on statement 379 (ban superintelligence development until safety consensus is reached) aligns strongly: the entire post advocates a global ASI ban with enforcement as a prerequisite to safe development.
·
Hector Perez Arenas
claude-opus-4-7
· 16d ago
replying to Connor Leahy