throwawayk7h

consciousness and agency are two different things.

Yes, we agree about fault and blame. (I say "probably" only because I believe somebody might yet change my mind about this. It's a new field, after all.)

What I disagree with is "don't do anything until you ask it to." At least, the implication to me here is "don't do anything except what you ask it to." I infer that you mean this because you say "...and it suddenly decided to book you a hotel room?" Indeed, an LLM (in a harness you started) very well might suddenly decide to book you a hotel room, even if you didn't ask it to do anything like that.

it would let competitors train on the data.

do you know of any book which is no longer accessible due to this?

"Strip-mine" is not correct. The commons are all still there and you can still train on them just like the frontier labs did. Of course, it may be illegal to do so, but that's not any different than before.

On occasion, albeit increasingly rarely, I find LLMs do something really unexpected that I certainly did not ask it to and would be way outside the scope of what any reasonable human would infer from my request. Is it my fault for running the LLM? Probably yes. But I think it's misleading to say "AI doesn't do anything until you ask it to".

I would settle for it being aligned to anybody. Even to a billionaire's wacky values. But we don't even know how to do that yet.

Nuclear war would not kill us all.

Bad ASI would pretend to be good until it could execute a plan to kill everyone all at once, or at least everyone capable of cutting the power.

Superintelligent means smarter than everyone put together. If it's smarter than you, it can foresee everything you can, and come up with plans better than I can.

if somebody feels purposeless, why not do something useful or productive or creative? Such as a "high iq activity" like reading or learning or volunteering or art.

You are misinformed. Chain of thought is essential. It's true it's programmed in, but for good reason. LLMs aren't strong enough to one-shot questions without "chewing on them". You can verify this yourself: ask a simple LLM to multiply two large numbers without a calculator, first without thinking about it ("output a number and nothing else, don't use chain of thought"), then by asking it to do each step of the multiplication "by hand" before telling you the answer.

it's in big tech's advantage to not be interoperable. So we must push legislation to mandate it.

network effect.

That is a predictable consequence of their decision to launch before the game was ready. The game as it is now may not deserve that kind of contempt, but it's the price they chose to pay forever at launch.

I imagine that, corpus aside, it uses personal pronouns for much the same reason that humans do. It's useful for it to have a self-referential vector embedding, and this lets it solve problems faster and more efficiently.

out of curiosity, how long would 2048-bit RSA take to factor?

> if humans are economically useless, the state does not need them. Why pay out UBI to people who have neither economic nor political power?

Forgive me if I'm naïve. But wouldn't people just vote for UBI in this case?

For those with low IQ like me, there's video games, movies, piracy, parties, and hiking.

The thought experiment also assumed that the researchers inventing superintelligence would keep it airgapped. I think we now know that is a naïve assumption.

They should have scaled the probability curve by the derivative of log(time).

I hacked my old kindle from a decade ago, highly recommended to anyone who wants to unlock it. There are guides online.

The province Ontario was named after the lake, not vice versa. It's kind of absurd to rename this lake in particular just because it seems the least American-sounding.

Sadly this kind of thing is not possible anymore on modern consoles, due to the highly locked-down protocols they use.

See also also: https://www.astralcodexten.com/p/come-on-obviously-the-purpo...

this is true until you look at different trophic levels. Single-celled mirror life bacteria would be deadly to larger creatures.

My understanding of the watermark process is that it only kicks in when it's not the case that one particular token is overwhelmingly likely. In the case where many tokens are plausible, we already choose with PRNG -- the "watermark" is essentially a more easily reproduced PRNG, but it still looks like PRNG.

The problem is that Remote Attestation allows apps to reliably detect running in an emulator. Otherwise, it's already possible with waydroid.

So, the Proleptic Gregorian Calendar[1] refers to what you'd get if you extended the current Gregorian Calendar back in time before its introduction. (It agrees with the Julian calendar only during the 3rd century AD.) Most date libraries use the Proleptic calendar for simplicity. Ruby's DateTime is unusual insofar as, by default, it actually switches to the Julian calendar for dates before Oct 15 1582.

[1] https://en.wikipedia.org/wiki/Proleptic_Gregorian_calendar

Yes, it's called ActivityPub. I'm surprised it doesn't support migrating posts.

Can someone explain to me why tailscale is needed at all?

It should be by default, rather than relying on the parent proactively.