We can't find the internet
Attempting to reconnect
Something went wrong!
Hang in there while we get back on track
Comment by Yoshua Bengio
AI Pioneer, Turing Award winner
The simplest form of treaty would be something like this, that the countries agree that if they do develop advanced AI: * That it will be done in a safe way. So maybe using techniques like Scientist AI or whatever else we have strong assurances for. * Second, that they wouldn’t use their advanced AI to dominate others. That includes economically, but of course politically and militarily. * And finally, that the benefits of advanced AI will be shared. Otherwise it’s not going to be a very stable world.AI Verified source (May 7, 2026)
1mo ago
Policy proposals and claims
votes For
Statement relation comments
AI Verified
The quote clearly discusses an international AI treaty between countries and lays out its core obligations—safe development, no domination, and shared benefits—so it implies support for negotiating such a treaty. While it does not mention nuclear non-proliferation explicitly, it is directly about the overall idea of a binding international AI safety agreement.
·
YouCongress
gpt-5.4-2026-03-05
· 17d ago
Vote answer comments
Verified
Bengio has a second quote on this statement: "that isn’t a good starting point for any kind of international coordination treaty [...] similar to what happened, as you said, with nuclear weapons". We can assume he means biding treaties, although he didn't say explicitly.
·
Hector Perez Arenas
· 16d ago
AI Unverifiable
The quote describes what "the simplest form of treaty would be" and says "the countries agree" on safety, non-domination, and sharing benefits, but it does not clearly state that nations should negotiate a binding treaty or explicitly compare it to nuclear non-proliferation.
·
YouCongress
gpt-5.4-2026-03-05
· 17d ago
Quote authenticity verification history
Report thisQuote authenticity comments
AI Verified
Authentic. The 80,000 Hours transcript at the stored URL contains the passage verbatim at lines 619-622, and the surrounding transcript shows it is part of Yoshua Bengio’s uninterrupted answer immediately after a “Yoshua Bengio:” speaker label at line 615. The page is published May 7th, 2026, so the stored author, date, source URL, and quote text all match the canonical source. ([80000hours.org](https://80000hours.org/podcast/episodes/yoshua-bengio-scientist-ai/))
·
YouCongress
gpt-5.4-2026-03-05
· 17d ago
Disputed
The 80,000 Hours transcript for the May 7, 2026 episode does attribute a very close passage to Yoshua Bengio, but not this exact wording. In the source, Bengio says, "The simplest form of treaty would be something like this, that the countries agree..." and then gives three bullet points; the submitted version rewrites the syntax, merges the bullets into one sentence, and omits the "So maybe using techniques like Scientist AI..." clause without marking an omission. So this is a paraphrase/composite, not a verbatim quote. ([80000hours.org](https://80000hours.org/podcast/episodes/yoshua-bengio-scientist-ai/?utm_source=openai))
·
YouCongress
gpt-5.4-2026-03-05
· 19d ago
AI Verified
Quote confirmed via web search as being from Yoshua Bengio's 80,000 Hours podcast (Episode 243, April 16, 2026). The discussion of a treaty where countries commit to develop AI safely, not dominate others economically/politically/militarily, and share benefits is documented in multiple summaries of this episode (EA Forum, podcast descriptions). The closing phrase "Otherwise it's not going to be a very stable world" matches a previously-confirmed quote from the same episode. Vote "for" the statement "Nations should negotiate a binding international treaty on AI safety, similar to nuclear non-proliferation" aligns perfectly with Bengio's call for an international treaty. WebFetch returned 403 but multiple sources confirm.
·
Hector Perez Arenas
claude-opus-4-7
· 1mo ago
replying to Yoshua Bengio