eru

28 ms·

661.
by eru 3mo ago
See the linked article: at least one sophisticated hacking attempt made the news. Of course, other ones might have happened in the dark. But it's fairly easy to image in hacking attempt like in the article, but with the additional st
662.
by eru 3mo ago
It would be cool (and scary), but also: there's largely no need for AIs to exfiltrate themselves. See https://en.wikipedia.org/wiki/Meme The thing that drove the AI here to do the intrusion came from a particular
663.
by eru 3mo ago
It's moot in practice, because there's additional sources of randomness from executing in parallel. Though in principle saving random seeds is a lot less hassle than keeping entire checkpoints around: your random seeds would fit o
664.
by eru 3mo ago
> That’s why developers should have as fast computers money can buy! I don't see the connection? Most local builds are done with debugging on and optimisation turned off anyway. And what you deliver to your customers is usually som
665.
by eru 3mo ago
Reproducing a stationary probability distribution that subsequent runs draw from is also a kind of reproduction. And presumably for ffmpeg you can fix the random seed? Tell me more about that design, please.
666.
Show HN: find Your misplaced iPad(github.com)
4 points by eru 3mo ago|1 comments
667.
by eru 3mo ago
> Should have minimal impact on the final outcome, [...] I share this expectation. But this is an interesting empiric question that deserves study; even if just to confirm what 'everyone knows'.
668.
by eru 3mo ago
Depends on your compiler. You could have a compiler that deliberately uses randomised algorithms. They are often faster and easier to understand and write. Though in practice you can get all the benefits of both determinism and (that kind
669.
by eru 3mo ago
> (As an aside, this is why the "open weights are not open source" thing is a complete misunderstanding. The weights themselves along with the documentation give you enough to fine tune the LLM. You can't rebuild it from s
670.
by eru 3mo ago
> You say "cultural fit", I say "corruption", let's blow this whole joint. It's not necessarily corruption. If there's no conflict between principal and agent, it's fine. Just like when I sent my
671.
by eru 3mo ago
It's a common joke, but it's not something you would find literally in the Google-internal official material. They ain't that stupid. The joke ain't Google specific, I might have picked it up working in finance. If you
672.
by eru 3mo ago
Oh, ok, that makes sense! I was just going on a tangential rant about DK.
673.
by eru 3mo ago
The 'Dunning-Kruger effect' of internet folklore is not what was in the original paper. And the original paper hasn't held up to well either: https://en.wikipedia.org/wiki/Dunning%E2%80%93Kruger_effect#.
674.
by eru 3mo ago
That's why we got training at Google that more or less told us not to put these kinds of things in writing. Remember: the e in e-mail stands for evidence.
675.
by eru 3mo ago
Why would it be a manual pass? There's only so many DRM schemes, and you can even use your previous generation AI to help with the breaking. > More importantly, it's also explicitly illegal. Agreed!
676.
by eru 3mo ago
In 1600 travel was slow, and safety pins hadn't been invented. But that doesn't mean safety pins sped up travel. Non-poisonous food is what economists call a 'normal good'. See https://en.wikipedia.org/w
677.
by eru 3mo ago
It's fairly easy (and cheap!) to run your own email server, but most people don't bother and just use gmail. And rightly so!
678.
by eru 3mo ago
> Pretty much anything that is still under copyright should be valuable only for its content, not for the physical medium it's printed on. Well, there are special collectors editions with the signature of the author and gold pages o
679.
by eru 3mo ago
There might be special rules around DRM that go beyond normal copyright?
680.
by eru 3mo ago
I agree. As for making the scans available: They could give a copy of the data once to some third party organisation that then seeds it in bittorrent or something like that. Basically, what I want to say is that this doesn't need to
681.
by eru 3mo ago
It's pretty expensive for the wildlife. Most rare books are rare because no one cared enough about them. Ie most rare books are rubbish.
682.
by eru 3mo ago
Empirically, probably not. We had lots and lots of content before copyright, and people seem to produce lots of content even in jurisdictions with weaker copyright.
683.
by eru 3mo ago
You can stick a full copy of the compiler and the zip binary into that archive, if you want to. The 'intron' can be arbitrarily large. You can add the complete schematics for your CPU. And the standard model of particle physics.
684.
by eru 3mo ago
I'm with Turing on this one.
685.
by eru 3mo ago
Determinism is interesting. It's mostly a question of 'deterministic in what context'. Eg the rules of chess are written as if the players have free choice. Similarly, algorithms are often analysed as if the antagonist that
686.
by eru 3mo ago
Yes, our current systems are likely still to primitive to be properly conscious. Probably. But I wouldn't rule out that ChatGPT 10 or Claude Fable 15 might get conscious.
687.
by eru 3mo ago
Well, thermodynamics is considered pretty fundamental, and the second law of thermodynamics is pretty much all about information. (But your link already touches that.) People also interpret the 'unitarity' of quantum mechanics to
688.
by eru 3mo ago
> Is it even possible that a complete simulation of the rules of this universe (let alone the content of it) could exist in this universe? Yes, it is possible! See how a Quine is constructed. https://en.wikipedia.org/wik
689.
by eru 3mo ago
> [...] but it presupposes that there is a theory of everything that can be used to perfectly simulate the universe, [...] LLMs are nice here: we don't need a theory of everything to simulate them. You can indeed use pebbles on the
690.
by eru 3mo ago
Yes, but that one has 2FA, and I might not travel with that second factor on me. Of course, they won't be amused.

More ›