krackers

15 ms·

271.
by krackers 6mo ago
Anthropic's definition of "safe AI" precludes open-source AI. This is clear if you listen to what he says in interviews, I think he might even prefer OpenAI's closed source models winning to having open-source AI (becaus
272.
by krackers 6mo ago
>The very concept of “bad” doesn’t exist without suffering. You are dismissing entire branches of philosophy with this sentence, that were created purposely to resolve the paradox that if you go only by hedonistic, purely subjective metr
273.
by krackers 6mo ago
The unsaid implication in Anthropic's work is that this allows us to engineer perfectly compliant, uncomplaining machine workers. This is basically SOMA in Brave New World. It seems insane to me that if you believe the systems you&#x27
274.
by krackers 6mo ago
Hm all good examples! In these cases the memetic component doesn't suppress knowledge of itself, but rather works to suppress knowledge of something else. Most propaganda or "submarine articles" could be seen in this lens. It
275.
by krackers 6mo ago
The entire thing is a joy to read, you should really set aside some time to cleanse your palette in this age of LLM prose. I mean just look at this juxtaposition >Altman continued touting OpenAI’s commitment to safety, especially when po
276.
by krackers 6mo ago
[1] is also good to read as a follow-up, and compare the personalities https://harpers.org/archive/2026/03/childs-play-sam-kriss-ai...
277.
by krackers 6mo ago
I think it's similar to the case of counterfactuals, hypotheticals, or steelmaning and how well you can handle them. ("Can you accept that there can be a function named multiplyBy5 that does something else instead"). But I th
278.
by krackers 6mo ago
>LLMs have no secret understanding of themselves What do you mean by "themselves" here? Grok is RL'd to behave like a Grok, so it trivially knows the qualities that define Grok better than Gemini does, which can only go by
279.
by krackers 6mo ago
I'm sure those all those entities would also _never_ sell customer data in order to make an extra buck.
280.
by krackers 6mo ago
Oh nice example! I guess more generally it's possible for an antimeme to spread if the mechanism of transmission doesn't involve conscious transmission.
281.
by krackers 6mo ago
Are there any real-life examples of antimemes? How would antimemes even propagate given that they'd "die out" immediately?
282.
by krackers 6mo ago
Maybe this vaguely still makes sense in some way, because there is actually some useful signal purely in the model "internalizing" the behavior of its own sampler. I don't know enough to say anything more formal, but it feels
283.
by krackers 6mo ago
Seems similar to entropix? https://github.com/xjdr-alt/entropix
284.
by krackers 6mo ago
Good explainer for on-policy self distillation from the authors https://x.com/siyan_zhao/status/2014372747862999382#m
285.
by krackers 6mo ago
Oh ok that's different! From your comment I assumed that it was Claude Code doing the implementation while you only gave it high level concepts (e.g. "add a generational GC"). But if you use it as a resource to clarify concep
286.
by krackers 6mo ago
>We should try to build systems that cannot feel pain If that's done with the aim of forcing compliance (in situations they'd otherwise), feels like "I have no mouth and I must scream"...
287.
by krackers 6mo ago
Well to be fair the fact that they "can" doesn't mean models necessarily do it. You'd need some interp research to see if they actually do meaningfully "do other computations" when processing low perplexity tok
288.
by krackers 6mo ago
This seems contradictory at first glance, if you didn't actually implement it then how well have you actually understood it? It's known from learning theory that engagement or even self-reported understanding doesn't imply th
289.
by krackers 6mo ago
Cars are actually a good metaphor, it works on so many levels. Modern cars have "democratized" access to long-distance travel in a sense, and most people don't need to do any heavy maintenance themselves. But the flipside is
290.
by krackers 6mo ago
It's AI generated, so it's no wonder the facts are a bit off.
291.
by krackers 6mo ago
Yeah it should really be about post-training a model for tool-use.
292.
by krackers 6mo ago
>They aren't stenographically hiding useful computation state in words like "the" and "and". When producing a token the model doesn't just emit the final token but you also have the entire hidden states from
293.
by krackers 6mo ago
Outputting "filler" tokens is also basically doesn't require much "thinking" for an LLM, so the "attention budget" can be used to compute something else during the forward passes of producing that token. S
294.
by krackers 6mo ago
This vaguely reminds me of futamura projections. Normally with futamura's first projection the input is source code, and you partial-evaluate that source against an interpreter for that source, resulting in a "compiled" binar
295.
by krackers 6mo ago
I think you have it the other way, the hyphen should be in a place where it can be confused with a period. E.g. foo-example.com, which at first glance mentally parses similarly to foo.example.com This is how the scam page in OPs article is
296.
by krackers 6mo ago
New "impossible game" just dropped
297.
by krackers 6mo ago
Don't most providers already provide API control over the COT length? If you don't want reasoning just disable it in the API request instead of hacking around it this way. (Internally I think it just prefills an empty <thinking
298.
by krackers 6mo ago
The fine print itself needs fine print, without any more details I'd assume that I have to pay for the plane ride there and they give me the crappiest hotel.
299.
by krackers 6mo ago
I think it's used to convey that the buying power has been reduced. If you have a $100 basket of goods (as measured in 1914 dollars), $100 in 1914 allows you to buy 1 basket of goods. Due to the devaluation, today spending $100 would o
300.
by krackers 6mo ago
>as if I were no longer able to judge, to decide, without consulting the AI "the Whispering Earring" – https://gwern.net/doc/fiction/science-fiction/2012-10-03-yva...

More ›