Zac Hatfield-Dodds has bet $10k to Ben Goldhaber's $1k that there will be no unpickable lock by 2027. This is related to claims that provably safe AI will be feasible.
Zac says:
"I win at the end of 2026, if there has not been a formally-verified design for a mechanical lock, OR the design does not verify it cannot be mechanically picked, OR less than three consistent physical instances have been manufactured. (e.g. a total of three including prototypes or other designs doesn't count)
You win if at the end of 2027, there have been credible and failed expert attempts to pick such a lock (e.g. an open challenge at Defcon). I win if there is a successful attempt."
See https://gsai.substack.com/p/august-2024-progress-in-guaranteed and https://www.lesswrong.com/posts/P8XcbnYi7ooB2KR2j/provably-safe-ai-worldview-and-projects#Ku3X4QDBSyZhrtxkM for more details.
This market will resolve YES if Ben wins the bet, NO if Zac wins, and will resolve N/A if the bet resolves as neutral or does not get resolved. I will rely on the opinion of the arbiter (@raemon) if there's a dispute.
@DavidFWatson I had to read the post multiple times troll I was certain I'm betting the way I meant to
@DavidFWatson Honestly it feels similar to halting problem to me and like there should be a proof that you cannot make a unpickable lock.
@Enlil it's a question of technical complexity vs human dexterity, I'd suspect it's eventually possible, but if it has not happened in all of human history, no reason to think it'll happen in a year or so.
Especially without a lot more money motivating it