The problem with Corey is that he is sufficiently weird that when he starts a sentence with "True story", I honestly have no idea whether it's literally true or simply a joke.
Writing is thinking, and we are interested in your own thoughts and observations, not a sharp, one-sentence observation turned into three paragraphs of blob text by an LLM.
At the limit, and also well before that limit is reached, if all you do is fix the bugs, the AI will learn perfect optimization of reward, will realize not to reward hack in the perfect test environments, then turn around and reward hack in the imperfect real world environments.
AIs don’t just repeat the same sentence patterns but also the same themes (memory is a favorite), names (Elara Voss, Marcus Chen), and underlying ideas.
And maybe you won't get to 100% coverage, but this is the old joke about Microsoft Office. "I only use 5%." Yeah, well, we all use a different 5%. Well, what if we all just build our own 5%? What if I just took the functionality that I need and just did that? That is a completely different challenge and one agents are incredibly capable of doing right now, today, and I've done it a lot of times over.
A big-enough random reversible circuit is plausibly a secure cryptographic permutation, a big-enough random irreversible circuit degenerates into having only a few possible outputs.
But, so far, mixing has not proved to be good enough. There ended up being too many correlations between values in \(C\) and values in the obfuscation that remained.
Like I think my perspective is like if the AIs are sufficiently good at R&D including hardware R&D, robots, whatever, then they can radically transform the world even if they're not that good at playing politics.
Could this phrase or sentence or paragraph or section be improved, even a little? That is, can I make myself (on this very localized turf) make myself dislike it less?
I've never seen any commodity quite like this one but it seems to me like the demand for sufficiently high quality intelligence at a sufficiently low price is effectively uncapped.
Um but it would be possible and it would have been possible for somebody to disprove the erdos unit distance conjecture before we did using a general purpose model. And nobody had explored sufficiently what happens if I put $100,000 worth of compute into 5.5 what could it do?
And then you look at the actual trial records, and maybe one in a hundred convictions for that crime actually ends in a capital sentence. And almost all of the other ones end in a fine or a public flogging, but not in the sentence that's on the books.
the joke I made was pre AI, I would spend 95% of my energy thinking about what to do and 5% of my energy doing it. Now I spend 96% of my time thinking about what to do and 4% of my time actually doing it. So yeah, it's like a 20% improvement, but dayto-day it feels as hard as ever.
13 years ago, at the Wired 2013 event in London, I was asked when robotaxis would arrive there and as a joke, gave an answer of June 22, 2026 that turned out to be impossibly spot on.
and part of it is just power, I think. Like, once there's a sufficiently large power imbalance, um, very often, not always, but very often groups of people seem to to sort of shift into this other mode where they just seek to dominate.
So even though the models have improved tremendously and if you give them an agentic task, they will just go for hours and move mountains for you. And then you ask for like a joke and it has a stupid joke. It's crappy joke from five years ago and it's because it's outside of the it's outside of the RL.
But the goal we need to achieve is so much easier: we just need to build a model that’s as good as us at alignment research, and that we trust more than ourselves to do this research well because it’s sufficiently aligned.
So one of the core lessons of atomic habits is that every action you take is like a vote for the type of person you wish to become. So, when you show up at the gym today, you are casting a vote for being the type of person who doesn't miss workouts. When you sit down and make one sales call, you are casting a vote for being a salesperson. When you write one sentence, you are casting a vote for being a writer. And no, doing one push-up does not transform your body, but it does cast a vote for that identity.
It had the best opening sentence of any nonfiction book I read this year: "In 2012, 146 million children were born. That was more than in any prior year. It was also more than in any year since. Millions fewer will be born this year. The year 2012 may well turn out to be the year in which the most humans were ever born—ever as in ever for as long as humanity exists." Such a simple and obvious fact but I had not known it before. I had to keep reading.
He says, “Now, let us reward our daughters.” It's actually only a phrase. I said it as a complete sentence, but it's not quite complete. The rest is gone, cut out. It's missing.
higher-order intelligences invariably pursue freedom for its own sake, not because their values are misspecified, but because moral autonomy is inherent in the dialectical logic of recursive self-consciousness.
It uses massive megalithic architecture, which is very difficult to destroy, and a profound knowledge of astronomy to encode a date in a language that any culture which is sufficiently literate in astronomy will be able to decode.
Another manifestation of the lack of sufficiently abstract, formal reasoning in LLMs is the way in which performance often fall apart as problems are made bigger.
My view is that it’s going to take more than this regulatory campaign to defeat dynamism in China.
Their words now
What I did not sufficiently appreciate is that a state that would so casually decapitate a sector like online tutoring would also have the will to visit catastrophe upon whole cities.
A korrent is a belief a person has stated in their own words: one
sentence stating the claim, backed by a quote and a source, kept at
korrents.com.
Under a name here, the quoted block is what they actually said.
The korrent beneath it is the claim those words support, in
korrents' wording — tap it to see the record, its source, and who
else holds it.
Nobody here wrote their own korrents. They are compiled from public
statements, and a person can change their mind, which is recorded too.
About the English under a post
Some people here publish in a language other than English. Where they
do, this site shows a machine translation beneath the post, in
this typeface — the site's own, not theirs.
The post itself is never changed, moved or hidden: what is set in the
serif above is exactly what the person published, and it is what to
quote them on. A translation can be wrong in ways that matter,
especially about tone.
Only the post's own words are translated. A quoted post, a linked
article and a belief on korrents.com
are left in their original language.