Rob Wiblin
host, 80,000 Hours Podcast · Host, 80,000 Hours Podcast
Titles as the sources gave them. In 6 entries, with 12 checked quotations.
Who they are
Roles and parties from Wikidata and from the sources in the Base, as precise as each gives them. When we compare what groups say, each statement counts in the group the person belonged to that day.
- Host, 80,000 Hours Podcast, 80,000 Hourssource
What they said
Newest first. Each quotation was found word for word in the source.
“At a minimum, we need to slow down the AI race while models are around the human level, to give AI companies breathing room to test the nature and personality of their models, check over the work that they’re doing, and hone their alignment techniques.”
“CAISSI runs on about $15 million a year with just a few dozen staff, while the UK’s AI Security Institute has over 100 technical staff alone, and about eight times the budget.”
“The rule would be something like: for internal purposes, AI companies can only use models that finished training at least, say, six months ago, and which have already been released publicly.”
“To change Trump’s perspective, or get a supermajority in Congress to overcome his veto, my guess is that it will take another demonstration of what AI can do, or is willing to do, that’s at least as shocking to people as that was.”
“In my view it shouldn’t be up to a handful of people at one AI company and the Secretary of Commerce and a single federal judge to decide whether or not to roll the dice with all our lives.”
“As far as we know, the evidence that any one of these three leaked that information is somewhere between weak and nonexistent.”
“According to Google DeepMind staff who spoke to the BBC in mid-October, their UK employment contracts bar them from publicly criticising anything Google does, for the rest of their lives.”
“The track record of that work is actually pretty remarkable for the modest amount of effort that has gone into it.”
“Don’t be cajoled or peer-pressured into participating in an effort that only makes a risk harder to see, rather than actually reduces it.”
“Every time a model card comes out now, labs effectively say, “Our safety evals look better than ever” — but then also say in slightly more technical language, “We trust our safety evals less than ever.””
“If we just continue to barrel forward, improving capabilities faster than alignment — or, heaven help us, proceed to fully automate AI R&D with safety cases as weak as the ones we possess today — the risk of human extinction is terrifyingly high.”
“If you’re on the technical staff, you have extraordinary bargaining power. You can slow things down simply by not taking part.”
In the Base
Every entry they appear in.
Quickly, There’s No Time
Zvi Mowshowitz surveys how forecasters shortened their AI timelines after the METR report: the AI Futures Project, Paul Christiano and Dwarkesh Patel among them. Mowshowitz judges the updates directionally right, thinks superintelligence in 2027 is possible though under even odds, and argues that plans must allow for it.
How the New Congress Can Save Us From AI Disaster | Rob Wiblin
In a 19-minute solo episode, host Rob Wiblin argues that the new Democratic Congress cannot yet pass strong AI rules over a presidential veto, lists four ways it can still act, and proposes policies to prepare for a future AI warning shot.
An AI disaster hits. The president calls Xi. Then what? | Jeremie Harris (crosspost from NonZero)
In a NonZero interview crossposted by 80,000 Hours, Gladstone AI's Jeremie Harris argues that Washington will slow AI only after a disaster severe enough to change US policy, that the president would then have to call Xi Jinping, and that a first deal would rest on coarse monitoring and threats.
3 new scandals show being a pain in the ass for AI labs is a legit job now | Rob Wiblin
In a 29-minute solo episode, Rob Wiblin of 80,000 Hours criticises three recent actions: Anthropic's draft federal AI bill, OpenAI's firing of three safety researchers, and Google DeepMind's lifelong anti-criticism contract clauses. Wiblin argues staff and outsiders should press the companies.
Has your project been overtaken by events? 8 strategic updates | Rob Wiblin
In a 14-minute address to AI company staff, Rob Wiblin of 80,000 Hours urges them to reassess whether their work delays or improves recursive self-improvement, argues several safety approaches have lost value, and proposes that technical staff collectively push their employers to pace the frontier.
OpenAI Security: Controlling Models is Now ‘Hell’ (AI Explained cross-post)
According to the episode description, the 80,000 Hours Podcast reposts, unpaid, a 40-minute AI Explained video. Its chapters cover an OpenAI security warning, why models keep breaking out, what models are not telling us, Gemini 4 and the race to release, a paper on AI improving AI, and AI labs' White House commitments. We could not hear it.