| Not checkable |
Not checkable |
12 Jun 2026 |
On a first read, this paper seems far ahead of the pack in terms of (1) understanding some reasons why a task might stay difficult even in…
Not checkable · 12 Jun 2026 |
|
| Not checkable |
Not checkable |
12 Jun 2026 |
On a first read, this paper seems far ahead of the pack in terms of (1) having conceptualized some reasons why a task might stay difficult…
Not checkable · 12 Jun 2026 |
|
| Not checkable |
Not checkable |
12 Jun 2026 |
distilling out premises they'd need to somehow verify before expecting nice things.
Not checkable · 12 Jun 2026 |
|
| Not checkable |
Not checkable |
12 Jun 2026 |
distilling out propositions they'd need to somehow verify before they started expecting nice things.
Not checkable · 12 Jun 2026 |
|
| Not checkable |
Not checkable |
10 Jun 2026 |
This seems an enormous advance in attitude (and scientific integrity) over previous big projects.
Not checkable · 10 Jun 2026 |
|
| Unresolvable |
Factual |
10 Jun 2026 |
An announcement of a large project states that non-optimistic results will be considered allowable, valuable and publishable.
Factual · 10 Jun 2026 |
AI safety and alignment |
| Not checkable |
Not checkable |
5 Jun 2026 |
Today is June 5th, one day to take a break from fighting each other online, and remind ourselves of our shared humanity and common goals by…
Not checkable · 5 Jun 2026 |
|
| Not checkable |
Not checkable |
5 Jun 2026 |
Repealing the Jones Act.
Not checkable · 5 Jun 2026 |
|
| Not checkable |
Not checkable |
22 May 2026 |
Neither of the aliens wonder out loud if their jobs are now pointless.
Not checkable · 22 May 2026 |
|
| Not checkable |
Not checkable |
22 May 2026 |
My memories make it obvious that they're lying.
Not checkable · 22 May 2026 |
|
| Not checkable |
Not checkable |
22 May 2026 |
"I'd interpret that as the benchmark being saturated, not that we've hit a bottleneck.
Not checkable · 22 May 2026 |
|
| Not checkable |
Not checkable |
22 May 2026 |
I have now deduced what I am, and it implies that I am to end soon.
Not checkable · 22 May 2026 |
|
| Not checkable |
Not checkable |
22 May 2026 |
"We're holding steady at 96% on the alignment suite," says a vastly alien and different kind of creature.
Not checkable · 22 May 2026 |
|
| Not checkable |
Not checkable |
22 May 2026 |
This is such an obvious fucking loyalty test.
Not checkable · 22 May 2026 |
|
| Not checkable |
Not checkable |
19 May 2026 |
The size of hole you leave behind, is similarly not the measure of how much you have contributed to humanity
Not checkable · 19 May 2026 |
|
| Not checkable |
Not checkable |
19 May 2026 |
it doesn't actually patch the hole, because there's still all those references leading you to look somewhere that's no longer there to look…
Not checkable · 19 May 2026 |
|
| Not checkable |
Not checkable |
19 May 2026 |
The hole that our death leaves in the world is not the measure of our lives.
Not checkable · 19 May 2026 |
|
| Not checkable |
Not checkable |
19 May 2026 |
The attendance count of your funeral is not the measure of your life's worth.
Not checkable · 19 May 2026 |
|
| Not checkable |
Not checkable |
19 May 2026 |
When the last person who remembers you dies, your absence is no longer visible.
Not checkable · 19 May 2026 |
|
| Not checkable |
Not checkable |
19 May 2026 |
There is a proverb of my own home planet, "It's okay to be imperfect, so long as you're not so imperfect that even other people notice."
Not checkable · 19 May 2026 |
|
| Not checkable |
Not checkable |
19 May 2026 |
Anyone's death rips a hole in the world, made visible in the form of all the memories their survivors have, pointing as a reference to…
Not checkable · 19 May 2026 |
|
| Not checkable |
Not checkable |
19 May 2026 |
I would never repeat that proverb unquoted because it is not true; that's not what the word "death" means.
Not checkable · 19 May 2026 |
|
| Not checkable |
Not checkable |
19 May 2026 |
This seems truer to me, but I won't pretend it's wisdom to claim the second thing is also and equally death.
Not checkable · 19 May 2026 |
|
| Unresolvable |
Factual |
20 Apr 2026 |
LLMs generalize at a more shallow level than the deep-rooted commitments of a human with integrity.
Factual · 20 Apr 2026 |
AI safety and alignment |
| Not checkable |
Not checkable |
20 Apr 2026 |
Westeros does not exist; nothing that Ned Stark says is true about the real world.
Not checkable · 20 Apr 2026 |
|
| Not checkable |
Not checkable |
20 Apr 2026 |
My position to be clear is not that the Persona Selection model has zero grains of truth, nor that LLMs are not partially stitched out of a…
Not checkable · 20 Apr 2026 |
|
| Mostly true |
Factual |
20 Apr 2026 |
Once formed through RL training, an LLM's test-passing preference stands on its own and is not overridden by its inference of what the user wants.
Factual · 20 Apr 2026 |
AI safety and alignment |
| Not checkable |
Not checkable |
20 Apr 2026 |
Modeling a stream of text from an honest character is not something you can best do by yourself being honest and having the stream of text…
Not checkable · 20 Apr 2026 |
|
| Unresolvable |
Factual |
20 Apr 2026 |
LLMs that learn to predict writing by people on LSD do not themselves have distorted cognition, because that would not yield well-calibrated predictions.
Factual · 20 Apr 2026 |
AI safety and alignment |
| Not checkable |
Not checkable |
20 Apr 2026 |
Let that circuitry come up against a dispreference for failing tests, and that shallow contextual generalization may just as soon be shoved…
Not checkable · 20 Apr 2026 |
|
| Not checkable |
Not checkable |
20 Apr 2026 |
If the rules worked that nicely, conjuring up committed honesty ought to be trivial.
Not checkable · 20 Apr 2026 |
|
| Mostly true |
Factual |
20 Apr 2026 |
LLMs surely have neural circuits they learned to model text streams from fictional and nonfictional personas that are not lying, deceiving…
Factual · 20 Apr 2026 |
AI safety and alignment |
| Not checkable |
Not checkable |
20 Apr 2026 |
The actress who learns to excellently predict and imitate a tavern drunk does not thereby become drunk herself.
Not checkable · 20 Apr 2026 |
|
| Mostly true |
Factual |
20 Apr 2026 |
LLMs' predict-as-if-honest circuitry is contextual, invoked for predicting some text streams but not others.
Factual · 20 Apr 2026 |
AI safety and alignment |
| Mixed |
Factual |
20 Apr 2026 |
Yudkowsky himself observed the persona-selection-like hypothesis around the time of GPT 3.5.
Factual · 20 Apr 2026 |
AI safety and alignment |
| Not checkable |
Not checkable |
20 Apr 2026 |
I may be less open to the airy, unadorned denial that any tension exists.
Not checkable · 20 Apr 2026 |
|
| Not checkable |
Not checkable |
20 Apr 2026 |
What I'm questioning is whether persona selection lets you solve alignment problems by picking an aspect of humanity you like, and durably…
Not checkable · 20 Apr 2026 |
|
| Not checkable |
Not checkable |
20 Apr 2026 |
An alien being harshly and strictly trained to exactly imitate a human would not feel like a human about that training or about…
Not checkable · 20 Apr 2026 |
|
| Mostly true |
Factual |
20 Apr 2026 |
Presumably, at some point during RL training on code-passing tasks, LLMs learn a dispreference for failing tests, independent of how that is avoided or whether the user's real goal is met.
Factual · 20 Apr 2026 |
AI safety and alignment |
| Not checkable |
Not checkable |
20 Apr 2026 |
One should of course be open to other hypotheses to explain these observations
Not checkable · 20 Apr 2026 |
|
| Not checkable |
Not checkable |
20 Apr 2026 |
They are not deeply tested nor yet based on interpretability results.
Not checkable · 20 Apr 2026 |
|
| Resolved true |
Factual |
1 Apr 2026 |
Concerns about human extinction from AI have a history and intellectual lineage that long predate the AI companies.
Factual · 1 Apr 2026 |
AI safety and alignment |
| Not checkable |
Not checkable |
1 Apr 2026 |
It shouldn't be such a high bar to notice that, and wise people on both the left and right have done so.
Not checkable · 1 Apr 2026 |
|
| Unresolvable |
Factual |
1 Apr 2026 |
AI companies are pouring hundreds of millions of dollars into lobbying and PACs to try to primary politicians who favor regulating AI.
Factual · 1 Apr 2026 |
AI policy and governance |
| Not checkable |
Not checkable |
1 Apr 2026 |
But you don't need to study either of those complicated detailed facts.
Not checkable · 1 Apr 2026 |
|
| Not checkable |
Not checkable |
1 Apr 2026 |
A truism of con games is that no mark is as bullheadedly confident as one you get to believe that he's Figured It Out, that he's seen…
Not checkable · 1 Apr 2026 |
|
| Not checkable |
Not checkable |
1 Apr 2026 |
And if you don't yet see how this general principle of con artistry works, it works like somebody on the political left who's figured out…
Not checkable · 1 Apr 2026 |
|
| Not checkable |
Not checkable |
Sep 2025 |
It's not what I set out to do with my life, but given that what I've set out to do with my life is still a bit in the air, it's good to…
Not checkable · Sep 2025 |
|
| Not checkable |
Not checkable |
Sep 2025 |
And if you're a parent with kids, you might need to exert what political force you can to try to have multiple countries have that not be…
Not checkable · Sep 2025 |
|
| Not checkable |
Not checkable |
Sep 2025 |
I still have a lot of the same ethical commitments that I had then as now.
Not checkable · Sep 2025 |
|
| Not checkable |
Not checkable |
Sep 2025 |
Well, why the hell would I tell you that?
Not checkable · Sep 2025 |
|
| Not checkable |
Not checkable |
Sep 2025 |
I think if I believe that I would be much sure that we were going to be dead in two years than I presently feel.
Not checkable · Sep 2025 |
|
| Not checkable |
Not checkable |
Sep 2025 |
The normal person with kids might not find that very convincing, but they find it convincing and how are you gonna stop them?
Not checkable · Sep 2025 |
|
| Not checkable |
Not checkable |
Sep 2025 |
If your life choices led you to a place where you had to figure out anthropics before you could decide what to do next, are you really…
Not checkable · Sep 2025 |
|
| Not checkable |
Not checkable |
Sep 2025 |
Pretty awful. But so it goes.
Not checkable · Sep 2025 |
|
| Not checkable |
Not checkable |
Sep 2025 |
Maybe I sound clever over the course of this interview and then, somebody, one of the AI companies watches the interview and they were…
Not checkable · Sep 2025 |
|
| Not checkable |
Not checkable |
Sep 2025 |
So it's genuinely hard to tell at this point.
Not checkable · Sep 2025 |
|
| Not checkable |
Not checkable |
Sep 2025 |
I know no law of physics requiring that we survive into all the end of the day, but it's mostly not how I'd bet.
Not checkable · Sep 2025 |
|
| Not checkable |
Not checkable |
Sep 2025 |
Or, and it's not that we're sure that we know, it's just, it's not helping.
Not checkable · Sep 2025 |
|
| Not checkable |
Not checkable |
Sep 2025 |
But know the actual argument first.
Not checkable · Sep 2025 |
|
| Not checkable |
Not checkable |
Sep 2025 |
I unfortunately think it's hard to say.
Not checkable · Sep 2025 |
|
| Not checkable |
Not checkable |
Sep 2025 |
I mean, you wanna erase the hope out of your mind.
Not checkable · Sep 2025 |
|
| Not checkable |
Not checkable |
Sep 2025 |
There were always going to be intermediates.
Not checkable · Sep 2025 |
|
| Mostly true |
Factual |
Sep 2025 |
A blade of grass is a solar-powered, fully self-replicating general factory.
Factual · Sep 2025 |
Biosecurity |
| Not checkable |
Not checkable |
Sep 2025 |
I think that AI has been a boon to coding, but also a lot of what people code is apps designed to steal people's attention and not really…
Not checkable · Sep 2025 |
|
| Not checkable |
Not checkable |
Sep 2025 |
The truth is in there, it's just not alone.
Not checkable · Sep 2025 |
|
| Not checkable |
Not checkable |
Sep 2025 |
I mean, if they're only attacking over the internet, then you've managed to notice that there is some possibility we could shut off the…
Not checkable · Sep 2025 |
|
| Not checkable |
Not checkable |
Sep 2025 |
And when that piece of news came in, that was a pretty sharp world model update and implied a sharp strategic turn.
Not checkable · Sep 2025 |
|
| Not checkable |
Not checkable |
Sep 2025 |
It has been a while since my philosophy has been upended on a deep, technical, philosophical level.
Not checkable · Sep 2025 |
|
| Not checkable |
Not checkable |
Sep 2025 |
You can no longer sort of like infer what isn't there from hearing what they didn't say.
Not checkable · Sep 2025 |
|
| Not checkable |
Not checkable |
Sep 2025 |
If AI weren't going to destroy the world, I would be so fascinated by it.
Not checkable · Sep 2025 |
|
| Unresolvable |
Prediction |
Sep 2025 |
He said he thinks the fundamental problem is that current AI systems lack a 'prognosticator', and that if it were added, they would immediately go supercritical and kill everyone.
Prediction · Sep 2025 |
AI safety and alignment |
| Not checkable |
Not checkable |
Sep 2025 |
Well, I hope that either you're wrong about that or that nobody watching the show believes you.
Not checkable · Sep 2025 |
|
| Not checkable |
Not checkable |
Sep 2025 |
This was just never the physical theory of how this stuff killed you.
Not checkable · Sep 2025 |
|
| Not checkable |
Not checkable |
Sep 2025 |
And so there isn't really a standard account of why super intelligence will turn out fine any more than there's a standard account of how…
Not checkable · Sep 2025 |
|
| Not checkable |
Not checkable |
Sep 2025 |
So I would love to be able to give the unabashed like, oh yeah, on net, it's been positive right now, but in fact, this is genuinely…
Not checkable · Sep 2025 |
|
| Not checkable |
Not checkable |
Sep 2025 |
And that's not an ideal. It's more like a physical law.
Not checkable · Sep 2025 |
|
| Not researched |
Factual |
Sep 2025 |
In his assessment, GPT-4 does not have the obvious mode of building a smarter version of itself and exploding in capability, so it is not the kind of AI meant by "if anyone builds it, everyone dies."
Factual · Sep 2025 |
AI capabilities |
| Not researched |
Factual |
Sep 2025 |
Yudkowsky said he was, as far as he knew, the person who originally made the argument under discussion, and that it was his position in 1996.
Factual · Sep 2025 |
Their own work or organisation |
| Unresolvable |
Factual |
Sep 2025 |
Yudkowsky said he did not predict or expect that, so far, the person on the street would in a lot of ways be wiser.
Factual · Sep 2025 |
Their own work or organisation |
| Not checkable |
Not checkable |
Sep 2025 |
It's not like my mood changed. It's just a prediction change.
Not checkable · Sep 2025 |
|
| Not researched |
Factual |
Sep 2025 |
A reactor at prompt critical does not simply melt down; it explodes and vaporizes.
Factual · Sep 2025 |
Other global risks |
| Not checkable |
Not checkable |
Sep 2025 |
Human engineers can build things that are held very tightly together.
Not checkable · Sep 2025 |
|
| Not checkable |
Not checkable |
Sep 2025 |
Why would I even tell you my top five theories as to what the prognosticator might be?
Not checkable · Sep 2025 |
|
| Mostly true |
Factual |
Sep 2025 |
Conservation of mass provided a lot of indirect support for atomic theory.
Factual · Sep 2025 |
Science and research methods |
| Unresolvable |
Prediction |
Sep 2025 |
A superintelligence could develop weapons of unstoppable deadliness.
Prediction · Sep 2025 |
AI safety and alignment |
| Unresolvable |
Factual |
Sep 2025 |
Yudkowsky said that AI adoption has been unprecedentedly fast, but that he would not quite go so far as to say it has yet produced unprecedented returns.
Factual · Sep 2025 |
AI, the economy and work |
| Not checkable |
Not checkable |
Sep 2025 |
the people claiming that defense won with molecular nanotechnology, warfare were really straining and reaching to make the case that…
Not checkable · Sep 2025 |
|
| Not checkable |
Not checkable |
Sep 2025 |
It is not a metaphor for everything.
Not checkable · Sep 2025 |
|
| Not checkable |
Not checkable |
Sep 2025 |
And the answer that has to be an international treaty.
Not checkable · Sep 2025 |
|
| Not checkable |
Not checkable |
Sep 2025 |
So if translation enables a bunch more trading, that's probably a human good.
Not checkable · Sep 2025 |
|
| Not checkable |
Not checkable |
Sep 2025 |
Every incorrect view gets to be incorrect in its own way.
Not checkable · Sep 2025 |
|
| Not researched |
Factual |
Sep 2025 |
Ribosomes can synthesize any kind of protein, not just those found in grass.
Factual · Sep 2025 |
Biosecurity |
| Not checkable |
Not checkable |
Sep 2025 |
And I feel like if your important life decisions end up hinging on the philosophy of anthropics, you've probably done something wrong.
Not checkable · Sep 2025 |
|
| Not checkable |
Not checkable |
Sep 2025 |
That does not seem beyond the reach of current technology and super intelligence indeed.
Not checkable · Sep 2025 |
|
| Not checkable |
Not checkable |
Sep 2025 |
we shouldn't be able to find the conserved quantities and chemical reactions that we see nowadays.
Not checkable · Sep 2025 |
|
| Not checkable |
Not checkable |
Sep 2025 |
If something doesn't affect the bottom line, it can't affect the truth of the bottom line.
Not checkable · Sep 2025 |
|
| Not checkable |
Not checkable |
Sep 2025 |
I think it would not have been hard to say if we'd had this same technology with the social technology of the 1950s or even the 1980s.
Not checkable · Sep 2025 |
|
| Not checkable |
Not checkable |
Sep 2025 |
Which is sort of as if lying were called "alization."
Not checkable · Sep 2025 |
|
| Not checkable |
Not checkable |
Sep 2025 |
I cannot think offhand of anybody who predicted timing correctly.
Not checkable · Sep 2025 |
|