
Paul Christiano – We could be 12 months away from AIs that can overthrow the government
Watch now (99 mins) | "It would be hard for humanity to recover."

AIs might overthrow the government within a year – Paul Christiano
Handing over humanity
Even the optimistic visions of a post-AGI future leave little room for human self-determination


Notes on Existential Risk from Artificial Superintelligence
Earlier this year I decided to take a few weeks to figure out what I think about the existential risk from Artificial Superintelligence (ASI xrisk). It turned out to be much more difficult than I thought. After several months of reading, thinking, and talking with people, what follows is a discussion of a few observations arising during this exploration, including: Three ASI xrisk persuasion paradoxes, which make it intrinsically difficult to present strong evidence either for or against ASI xrisk. The lack of such compelling evidence is part of the reason there is such strong disagreement about ASI xrisk, with people often (understandably) relying instead on prior beliefs, self-interest, and tribal reasoning to decide their opinions. The alignment dilemma: should someone concerned with xrisk contribute to concrete alignment work, since it's the only way we can hope to build safe systems; or should they refuse to do such work, as contributing to accelerating a bad outcome? Part of a broader discussion of the accelerationist character of much AI alignment work, so capabilities / alignment is a false dichotomy. The doomsday question: are there recipes for ruin -- simple, easily executed, immensely destructive recipes that could end humanity, or wreak catastrophic world-changing damage? What bottlenecks are there on ASI speeding up scientific discovery? And, in particular: is it possible for ASI to discover new levels of emergent phenomena, latent in existing theories?
Q3.5 Timelines Update: Looks Like the Intelligence Explosion Will Happen in 2027 After All
We wanted to get this post out quickly given the urgency of the topic.

Independent Assessment of AI R&D Acceleration at Anthropic in October 2026
METR's high-access assessment of the degree to which AI is accelerating AI research and development at Anthropic.

Covert underperformance on safety research tasks · OpenAI Alignment
Research on aligning AI with human values and intent, and reports documenting model failures.
AI is dramatically accelerating AI research right now
If acceleration compounds, the naive extrapolation implies superintelligence within a year

The risks of advanced AI
Could working on AI risks be the highest-impact career choice today? Explore why AI may trigger rapid, dramatic societal change — and what you can do about it.



When "yang" goes wrong — LessWrong
On the connection between deep atheism and seeking control.
Moloch Hasn’t Won
This post begins the Immoral Mazes sequence. See introduction for an overview of the plan. Before we get to the mazes, we need some background first. Meditations on Moloch Consider Scott Alexander&…

Problems of Proliferation — LessWrong
Scientific progress and its perils. TLDR: The vulnerable world hypothesis is likely correct. Once enormous amounts of cognitive labor start getting a…
Here Are a Few of the Ways AI Could In Fact Kill Everyone
It has become fashionable to ask “doomers” to “get specific” about their dooming. Well, aiight then.


a post I've been wanting to make ever since the "METRSIPDATE" (metr rsi update) (don't quote this) is it seems like the futures have now condensed down to 2, call them high- and low-crazy. but no-crazy is sort of off the table at this point, in a way it wasn't in maybe 2022-2023.
Joshua Shew
I'm still making life plans measured in years that assume the way the economy is organized stays ~the same :/ This feels increasingly incongruous with my beliefs about how AI will change the world in the coming months/years. "Business as usual" seems an unlikely path.