fenomas

30 ms·

421.
by fenomas 4y ago
I don't think that premise stands. If I ask ChatGPT to write a poem about Sherlock Holmes fighting a mermaid it will happily do so, and I think most users would agree that it'd be a failure mode if it refused. Isn't it more t
422.
by fenomas 4y ago
It's so weird to me when apparently tech-savvy people say ChatGPT "lies", or call it a bullshit generator and so on. I mean - when you ask StableDiffusion to draw a dog astronaut, everyone gets that the image it returns is ma
423.
by fenomas 4y ago
No, TFA was a little vague but there's no collision. The kanji for mackerel 鯖 is a standalone word/idea, whose graphical representation happens to consist of the characters for "blue" and "fish" next to each ot
424.
by fenomas 4y ago
Who are you responding to? TFA doesn't discuss learning methods at all, let alone "promoting the rote-learning of Kanji". Is it just the mention of WaniKani? Personally I don't like WK either, but the author barely menti
425.
by fenomas 4y ago
> is a very different ball game than "our model understands the rules behind this sequence". The authors state plainly what they mean by "understands the rules". That the phrase could also mean other things is no reas
426.
by fenomas 4y ago
The crow never analyzes any board - that's the point of the analogy. It only hears a sequence of tokens, and it learns to predict future tokens, and TFA sets out to answer whether it does that by "surface statistics" (i.e.
427.
by fenomas 4y ago
> if it made no mistakes it would offer strong evidence I don't see how this has anything to do with the article. TFA's authors weren't trying to make the GPT "know the rules" of Othello, or anything similar. The
428.
by fenomas 4y ago
Re: TFA's complaints of blandness, and the several comments here about the narrator not being Watson, I'm always confused by how people expect ChatGPT to follow constraints it wasn't given. I guess it's a testament to ho
429.
by fenomas 4y ago
It's fairly bonkers how critics will say that ChatGPT or Lamda are just glorified predictive text in one breath, and then call them "arrogant" or "overconfident" in the next.
430.
by fenomas 4y ago
It seems to me that questions like this show how people misunderstand generative AIs. If you ask DALL-E to draw the Mona Lisa, it won't give you a pixel-perfect copy. It will give you *a* Mona Lisa - which won't be the Mona Lisa
431.
by fenomas 4y ago
I've always wished there was a social network that applied some sort of "I am interested in [person] by [amount]%" premise, but recursively. E.g., suppose I tell the network that I'm 80% interested in William Gibson, and
432.
by fenomas 4y ago
I strongly agree. ChatGPT ultimately tries to give you what you ask for - so if certain prompts cause it to give "leftist" (by whatever definition) answers, then one might reasonably conclude that something about those prompts mak
433.
by fenomas 4y ago
(When I replied to it, Toomim's post was wholly different and linked to a Ben Shapiro video about purported leftist bias in ChatGPT.)
434.
by fenomas 4y ago
My reply was to GP's assertions about which questions GPT will and won't answer, not to TFA's claims about political compasses. > Ben Shapiro ...really?
435.
by fenomas 4y ago
This sounds like a line being fitted to a single data point. Anyone who's played with ChatGPT a lot will have seen it sometimes refuse to answer wholly innocuous prompts, while other times it happily answers things it probably shouldn&
436.
by fenomas 4y ago
It runs on the inputs you give it, not the ones inside your head. If you don't specify what kind of notation you want, then colored-in notes is a perfectly reasonable notation for it to suggest. If you want something else, you can lite
437.
by fenomas 4y ago
What I said: that outputting a poem structured around having end rhymes for rhyme made-up words was an obvious counterexample to the claim that the system is just regurgitating its training data. What I did not say: that I was amazed becaus
438.
by fenomas 4y ago
> You asked it for a novel musical notation, and it gave you Newp - I asked it for "a visually unusual musical notation that isn't just adding extra markup to traditional staff notation", or words to that effect, and it su
439.
by fenomas 4y ago
It was an example of GPT offering a notation unrelated to anything mentioned in the prompt. I wasn't suggesting we all adopt it as our new notation. This whole comment page feels like a chain of "it clearly can't do X" -
440.
by fenomas 4y ago
> You're amazed because [stuff nobody said]? Comments like this add nothing to the conversation, please consider not posting them.
441.
by fenomas 4y ago
Heh. Out of curiosity I just tried what you're asking for, and ChatGPT suggested a color-based notation where parts for multiple instruments are displayed together in different colors, so that each musician can focus on their own part
442.
by fenomas 4y ago
I didn't claim the poem was interesting, I gave it as an example of something which could not possibly have been regurgitated straight out of the training set, which it clearly was. Also: it was a 4- stanza poem, three rhyming couplet
443.
by fenomas 4y ago
Constructing a novel, genuinely humorous poem that incorporates given inputs isn't very interesting because... I had to supply the inputs? Sorry, but I don't follow.
444.
by fenomas 4y ago
> regurgitate existing works The other night it regurgitated a humorous four-stanza poem for my D&D group, incorporating (and even rhyming ) several of the player character names. Given that the names were made-up and unique, and ha
445.
by fenomas 4y ago
Some months ago I did the same thing as TFA - took a look at Vue3 and decided to change my main project's UI from Vue2 to svelte. So far I've been really happy with it. In my case the big motivator was that with Vue, whenever I wo
446.
by fenomas 4y ago
The context can be found here: > https://en.wikipedia.org/wiki/Likelihood_principle#The_voltm... Specifically if you read the "Example 2" section above that blurb, and then read the "throwback" b
447.
by fenomas 4y ago
> Perhaps a closer analogy would be that the new boss rolled in and threw away 80% of the fire extinguishers. If we're deep-diving it'd be closer to say that he rolled in and sold 80% of the stuff , largely sight-unseen, and i
448.
by fenomas 4y ago
To me the analogy here is that the new boss rolled in and sold all the fire extinguishers. That by itself doesn't set the building on fire - it doesn't even increase the chances of a fire occurring on any given day. But when one d
449.
by fenomas 4y ago
I see the issue - it looks like it could be a "～って(いうのは)いいぜ" kind of construction. But in that case the preceding bit wouldn't be te form, it would be "動きやすいっていいぜ" or whatever. Rather, the っ here is just an affectat
450.
by fenomas 4y ago
TFA's version is right. "Shorts" is the topic word (the topic particle after it is implied), and the "easy to move" and "good" parts both modify "shorts". So robotically, "speaking of shorts

More ›