kogasa
@kogasa@programming.dev
- Comment on I wish I was as bold as these authors. 1 day ago:
It has access to a python interpreter and can use that to do math, but it shows you that this is happening, and it did not when i asked it.
That’s not what I meant.
You have access to a dictionary, that doesn’t prove you’re incapable of spelling simple words on your own, like goddamn people what’s with the hate boners for ai around here
??? You just don’t understand the difference between a LLM and a chat application using many different tools.
- Comment on I wish I was as bold as these authors. 1 day ago:
ChatGPT uses auxiliary models to perform certain tasks like basic math and programming. Your explanation about plausibility is simply wrong.
- Comment on I wish I was as bold as these authors. 2 days ago:
If you fine tune a LLM on math equations, odds are it won’t actually learn how to reliably solve novel problems. Just the same as it won’t become a subject matter expert on any topic, but it’s a lot harder to write simple math that “looks, but is not, correct” than it is to waffle vaguely about a topic. The idea of a LLM creating a robust model of the semantics of the text it’s trained on is, at face value, plausible; it just doesn’t seem to actually happen in practice.
- Comment on Deadrop developer Midnight Society cuts ties with Dr Disrespect following new Twitch ban allegations 5 days ago:
Well, we knew he was a shitbag beforehand, so that’s not really what’s in question
- Comment on Evidence 5 days ago:
I’m not a physicist, I don’t know one way or another. But it’s possible that there’s a leading explanation for the formation of the universe based on a mathematical model that predicts exactly one big bang.
- Comment on Evidence 6 days ago:
Based on the comment you’re replying to, I assume they would say "no, nothing materialized from nothing because there wasn’t a ‘before’ in which nothing could have existed’
- Comment on Cats 6 days ago:
It wouldn’t have been published, and he’s only relatively famous if you’re a topologist, but it was Charlie Frohman. Not that it must carry the same weight for you, but I value his insight highly, even if it’s just a quip.
- Comment on Science memes 1 week ago:
Yes, but it proves that termwise comparison with the harmonic series isn’t sufficient to tell if a series diverges.
- Comment on Science memes 1 week ago:
The assumption is that the size decreases geometrically, which is reasonable for this kind of self similarity. You can’t just say “less than harmonic” though, I mean 1/(2n) is “slower”.
- Comment on Cats 1 week ago:
Quoting a relatively famous mathematician, linear algebra is one of the few branches of math we’ve really truly understood. It’s very, very well behaved
- Comment on Irrational 1 week ago:
Google it? Axiomatic definition, dedekind cuts, cauchy sequences are the 3 typical ones and are provably equivalent.
- Comment on Irrational 1 week ago:
I’m fully aware of the definitions. I didn’t say the definition of irrationals was wrong. I said the definition of the reals is wrong. The statement about quantum mechanics is so vague as to be meaningless.
- Comment on Irrational 1 week ago:
That is not a definition of the real numbers, quantum physics says no such thing, and even if it did the conclusion is wrong
- Comment on Know who is king 2 weeks ago:
Only if you’re trying to get a numerical point evaluation. For example, one can use Fourier series to represent complex signals in terms of sine waves, and then reproduce the sine waves with hardware to reproduce the original signal.
- Comment on Tech Bros Invented Trains And It Broke Me - YouTube 2 weeks ago:
Would not be surprised if it’s getting confused about a wallet sponsorship
- Comment on Long Cow is coming 2 weeks ago:
We aren’t trying to establish that neurons are conscious. The thought experiment presupposes that there is a consciousness, something capable of understanding, in the room. But there is no understanding because of the circumstances of the room. This demonstrates that the appearance of understanding cannot confirm the presence of understanding. The thought experiment can’t be formulated without a prior concept of what it means for a human consciousness to understand something, so I’m not sure it makes sense to say a human mind “is a Chinese room.” Anyway, the fact that a human mind can understand anything is established by completely different lines of thought.
- Comment on Long Cow is coming 2 weeks ago:
This fails to engage with the thought experiment. The question isn’t if “the room is fluent in Chinese.” It is whether the machine learning model is actually comparable to the person in the room, executing program instructions to turn input into output without ever understanding anything about the input or output.
- Comment on Keep your friends close, keep your enemies closer 5 weeks ago:
I’m not suggesting that. It just looks like he did in fact choose to have that hair color.
- Comment on Keep your friends close, keep your enemies closer 5 weeks ago:
It looks dyed
- Comment on The future is here 5 weeks ago:
That’s not necessarily wrong, but not the big explaining factor here I think. The technological challenges behind aligning ML models with factual reality aren’t solved, so it’s not an engineering decision. It’s more that AI is remarkably easy to market as being more capable than it is
- Comment on LPT Do it. 5 weeks ago:
Typst is pretty functional
- Comment on Limericks 1 month ago:
It’s common to just use “log” in pure math, since nobody uses the base 10 log.
- Comment on Zero to hero 1 month ago:
I’m not.
- Comment on Zero to hero 1 month ago:
But not all mathematicians are logicians.
- Comment on Zero to hero 1 month ago:
It is true.
- Comment on Zero to hero 1 month ago:
Of course they’re considered equally viable conventions, it’s just that one is prevalent among Americans and the other isn’t.
- Comment on Zero to hero 1 month ago:
I did say mathematician, not logician.
- Comment on Zero to hero 1 month ago:
There can’t really be an argument either way. It’s just a matter of convention. “Natural” is just a name, it’s not meant to imply that 1 is somehow more fundamental than -1, so arguing that 0 is “natural” is beside the point
- Comment on Zero to hero 1 month ago:
Ehh, among American academic mathematicians, including 0 is the fringe position. It’s not a “debate,” it’s just a different convention. There are numerous ISO standards which would be highly unusual in American academia.
- Comment on 🔔 SHAME 🔔 1 month ago:
Limits at infinity are one thing, but infinite ordinals are meaningfully used in set theory and logic