drdeca

39 ms·

511.
by drdeca 2y ago
A few thoughts this brings to mind: 1) this reminds me of how I was thinking about “what if you tried to modify an n-gram model to incorporate a poor-man’s approximation of the copying heads found in transformer models? (one which is also b
512.
by drdeca 2y ago
If X runs in polynomial time, then there is of course some X' which produces the same outputs on the same inputs, and provably (in PA, I'm pretty sure, certainly in ZFC) runs in polynomial time. So, if there exists an algorithm X
513.
by drdeca 2y ago
What if you had an algorithm which could e.g. solve 3SAT in polynomial time, but couldn’t be proven (in ZFC) to always produce the correct answer?
514.
by drdeca 2y ago
Not just valid, but does a not-nonsense task. I think in this case, there is a coincidence on both sides? Like, the language or the painting could have been different such that the painting would run, but what it would do wouldn't be
515.
by drdeca 2y ago
The issue I see with the "can't express at all" view, is that, if it can't be expressed, then how do newborns learn to speak the language? If from some sequence of sense perceptions, a child can learn to associate some w
516.
by drdeca 2y ago
Well, I wouldn't say we are "in agreement" on this. The reason I commented initially just on the number of nines, is because that was a clear cut thing, which I thought kind of amusing. Certainly I agree that one cannot reall
517.
by drdeca 2y ago
I suppose, but, not on the basis of "they are granted the authority to do this thing, but will not abuse it", so much as, "it is costly for them to do this thing, and it is unlikely to be worth it to them" ?
518.
by drdeca 2y ago
Well, if the action produces "more utility". Lives is one measure, but, while "lives saved - lives lost" will probably play a role in most ways of defining "utility", I don't think most utilitarians would
519.
by drdeca 2y ago
No I mean, I gave the maximum number of nines that you could possibly be justified in listing. The maximum number of nines, would be in the case that I described: in which you meet every other person on earth, but only one of them will you
520.
by drdeca 2y ago
that is way too many nines. If in my life, I saw one person, and then saw every other person on earth, and then saw the first person again, and then died, then, after seeing each person on earth once, then the fraction that I would never
521.
by drdeca 3y ago
You said “ > It was proven with math” But that is not what I said! Do not put words in my mouth. I didn’t say that the math “proved” it. I was saying that the molasses analogy did not matter to what they did, and that the math, not the m
522.
by drdeca 3y ago
The thing you are asking for, the " why does mass cause spacetime to bend" seems... perhaps impossible to answer? Or like, I don't see why there should necessarily be anything that satisfies what seems to be your reason for
523.
by drdeca 3y ago
This doesn't appear to fit with other quantities like say, electric charge, or lepton number, or whatever. Like, there's more to matter than "how much is there here".
524.
by drdeca 3y ago
> Thought it was called "The God Particle" because it set out to explain "how particles got their mass" "[...] Why God Particle? Two reasons. One, the publisher wouldn't let us call it the Goddamn Particle [
525.
by drdeca 3y ago
What about astronomy and cosmology?
526.
by drdeca 3y ago
Perhaps you have misunderstood what the people you are talking about, mean? Or, if not, perhaps you are conflating what they mean with something else? Something doesn’t need to have had a subjective experience of the world in order to act a
527.
by drdeca 3y ago
I think they mean you can't, "mute the ad, and scroll down and read the comments (no longer seeing the ad, because one is scrolled down and reading the comments)", not because one can't mute the ad, but because scrolling
528.
by drdeca 3y ago
What? No, it's that the comments are now on the side, and that there's a bunch of video thumbnails where the comments used to be
529.
by drdeca 3y ago
Maybe they do now, but at one point I had a “Python2IDE” application on my iphone. (Well, technically it is still on my list of installed applications, but it isn’t compatible with the version of the OS I have installed now. It says it need
530.
by drdeca 3y ago
A finite state machine can control the head of a Turing machine (as, that is what controls the head of a Turing machine). A transformer could be trained to implement a particular finite state machine (using a fixed length context window). O
531.
by drdeca 3y ago
While of course one can use whatever orthonormal basis one likes, and so in particular one can use the one shown in the article, uh, I’m not sure what statement that others might believe, the article is contradicting? The Fourier transform
532.
by drdeca 3y ago
People are allowed to have preferences.
533.
by drdeca 3y ago
"chauvinism" ? What do you mean by that? Wikipedia describes "chauvinism" as "the unreasonable belief in the superiority or dominance of one's own group or people". Now, if there were silicon-based persons
534.
by drdeca 3y ago
As I use the terms, the dual numbers are a different thing from non-standard analysis. Non-standard analysis, as I understand the term, uses non-standard models of the real numbers, and its infinitesimals do not satisfy h^2 = 0. In non-stan
535.
by drdeca 3y ago
I don't know why you keep calling it "magic". Whether or not pytorch uses them, they aren't magic, neither in the derogatory sense nor in the praise sense.
536.
by drdeca 3y ago
> People keep calling it "next next token predictors", but clearly there is something more going on I think this depends what you mean by "something more going on". Now, if someone says that it is "just" &
537.
by drdeca 3y ago
I think there are still open questions about this that are worth asking. It is clear enough that following gradients of a bounded differentiable function can bring you to a local minimum of the function (unless I guess if there’s a path tha
538.
by drdeca 3y ago
Thanks! But, quadratic time complexity (which I had heard that attention requires, which doesn't surprise me) is not exponential time. I acknowledge that they related this to the assumption of the SETH (which surprises me a little). Bu
539.
by drdeca 3y ago
All of those before the simulation, seem like fairly useful metaphors, that describe something real about the world. Indeed, there are many things about the world that go in cycles. Indeed, much of how the world behaves can be seen as actin
540.
by drdeca 3y ago
> isn't this 'everything is identical' model fundamentally incompatible with quantum mechanics? No, quite the opposite, actually. Some particles being identical is core to many quantum mechanical ideas. The distinction bet

More ›