Recent submissions

Comment · 22 days ago
I missed "compiling shaders"
Comment · 1 month ago
Cough we're still waiting for those alignment dollars to really kick-in!!1
Comment · 1 month ago
Comment · 1 month ago
That doesn't mean you can hit it too. The only case which I remember was well documented was for a turret gun "controlled by satellite" hit the passenger of a standing vehicle from the back of a truck in front of it (nuclear scientist Mohsen Fakhrizadeh, killed in Tehran in 2020 while his wife, sitting immediately beside him, was unharmed). Also those reports were pretty questionable at the time, there were multiple competing narratives, and I'm not sure the satellite gun story was ever independently verified.
Comment · 1 month ago
Same!
Comment · 1 month ago
Comment · 1 month ago
Worthless is a pretty good description IMO in the context of what Lean is trying to achieve: "enable correct, maintainable, and formally verified code". Tens of millions of lines of LLM vomit may be many things, but it often turns out to not be correct and certainly not maintainable. Formally verified remains as a thin fig leaf covering the uncomfortable truth that formal methods only provide assurances under assumptions (your toolchain, libraries, compiler, OS, and hardware are "correct" and don't expose some exploitable flaw).

It doesn't mean that it cannot improve over time, maybe the proof can be "minified" to a state where human reviewers are able to comprehend it; but as it stands there isn't really much insight or confidence to be gained from the artifact itself.

Comment · 2 months ago
Right, exactly that was what I meant. Thanks for clarifying. I remember being surprised by this when I heard it on NPR a few weeks ago.
Comment · 2 months ago
Comment · 2 months ago
The implicit "men born outside the US are not men" is arguable worse.
Comment · 2 months ago
In the first case they (the US govmnt) could hold him (the citizen) in contempt (in a cell) indefinitely.
Comment · 2 months ago
While I tend to agree on the overall sentiment, I think this rebuke is inaccurate. Some of these "reasoning" models are trained using "Chain-of-Thought" where the model is presented explicit, intermediate reasoning steps (either by a human or some automation) that supposedly get it closer to the correct answer. These intermediate steps are what was originally called "thinking traces" - not what the model produces to mimic them.

But yes, anthropomorphizing model outputs leads to worse outcomes.

Comment · 2 months ago
Well, one obvious question is if the simulated brain experiences pain, or more generally can it suffer? And if so, how would we tell.
Comment · 2 months ago
But Matthew, think of the kids!!1

Honestly though, framing this as a "tech issue" doesn't help IMHO, it just muddies the water. Ever since RSA was invented privacy has been about educating people on how to use it effectively and _why they should care_. If voters now are choosing authoritarianism over democracy and individual freedom, I think we have to face the reality that after almost 50 years of fighting battle after battle on the technology front, we've largely lost the war on the home front in this regard.

Comment · 2 months ago
Yeah. This is basically why we need to "accelerate"..

Live from the official Hacker News API

A modern reader for Hacker News · SvelteKit + Skeleton