gjm11

19 ms·

181.
by gjm11 1y ago
My tests of GPT-5 were using the online chat version. Of course, I only ran it once; I can't at all rule out the possibility that sometimes it gets it wrong. But, again, the same is true of humans.
182.
by gjm11 1y ago
The fact that an LLM is running inside an ordinary computer does not mean that it gets to use all the abilities of that computer. They do not have megabytes of scratch space merely because the computer has a lot of memory. They do have so
183.
by gjm11 1y ago
I have a PhD, in mathematics, from a top university. If you give me, say, 100 10-digit numbers to add up and tell me to do the job in my head then I will probably get the answer wrong. Of course, if you give me 100 10-digit numbers to add
184.
by gjm11 1y ago
> GPT-5 and Gemini 2.5 still cannot compute an arbitrary length sum of whole number without a calculator. Neither can many humans, including some very smart ones. Even those who can will usually choose to use a calculator (or spreadsheet
185.
by gjm11 1y ago
Only to criticism of the form "X can never ...", and some such criticism richly deserves to be ignored. ( Sometimes that sort of criticism is spot on. If someone says they've got a brilliant new design for a perpetual motion
186.
by gjm11 1y ago
The heuristic doesn't work quite so well when applied to the actual original version of that line.
187.
by gjm11 1y ago
If you prefer "accusation" to "attack", that's fine with me. I take it that the actual meaning in this context is "something it's reasonable to respond to by defending oneself", since velcrovan was
188.
by gjm11 1y ago
> the ability to effectively blacklist people from a weird, unpleasant, mostly pointless, but highly remunerative niche labor market providing an expedited path to US citizenship Huh?
189.
by gjm11 1y ago
"Cet animal est très méchant, Quand on l'attaque il se défend." Saying in public that someone "replied with a bunch of very unconstructive crap-on criticism" and "[m]aybe he's different in person but based
190.
by gjm11 1y ago
In the few games I tried, it always ended up that way. You start with a few prime numbers (if you're lucky). Each time you click on one, it gets replaced with another number that usually isn't prime. After a short while, there are
191.
by gjm11 1y ago
I don't know whether it's different in the US, but here in the UK it's pretty common for teachers to teach part or all of an actual lesson as part of the application process. Experienced teachers as well as new graduates.
192.
by gjm11 1y ago
"Inspiration", perhaps.
193.
by gjm11 1y ago
I find this a perplexing comment in view of the fact that almost all of the linked article is in fact about how the author wrote his own date parsing library; the "never do it" bit is just a couple of lines at the start and so f
194.
by gjm11 1y ago
If I'm understanding your logic correctly, I think it's flawed. It seems like you're saying: if you choose to work for a startup rather than a bigger company, it must be because you think their equity is valuable, so you shou
195.
by gjm11 1y ago
An oxymoron is a word or phrase that contradicts itself (the word comes from Greek bits meaning "sharp" and "soft"). It doesn't make any sense at all to say that Python is an oxymoron. Python-the-language isn't
196.
by gjm11 1y ago
Neat. Is there any way to search the highlights? ("What are the most noteworthy things that have been said on HN about ...?" "I remember reading an excellent comment that had something to do with ..., but all I can remember i
197.
by gjm11 1y ago
It isn't clear to me what exactly OP means by "industrial" but it seems possible that they might not consider it to apply to TXR.
198.
by gjm11 1y ago
Your Nakamura example is from 2008. That's 17 years ago. The machines have improved a lot since then, hardware and software both. I've seen Nakamura beat strong-but-still-limited bots playing "anti-computer chess" but I
199.
by gjm11 1y ago
Human IMO contestants are also trained specifically on IMO problems.
200.
by gjm11 1y ago
My understanding is that they did , but don't any more; it's no longer true that humans understand enough things about chess better than computers for the human/computer collaboration to contribute anything over just using t
201.
by gjm11 1y ago
What is the purpose of this comment? It seems to be taken directly from the article, but it doesn't begin at the beginning or end at the end or shed any light on the part of the title most likely to be puzzling ("beyond arithmetic
202.
by gjm11 1y ago
> Hunger worldwide has been getting worse for the last quarter century or so. That doesn't appear to be true. E.g., following links from the WHO page you cite gets me to https://openknowledge.fao.org/server/api&
203.
by gjm11 1y ago
I'm not missing the point, I'm disagareeing with it. I am saying that the convention that if you say 20% then you are assumed to have an error margin of 5%, while if you say 19% you are assumed to have an error margin of 1%, i
204.
by gjm11 1y ago
It doesn't look like any exfiltration is happening? Although they use the word "telemetry", unless I'm misunderstanding the information is just stored locally on the device. (Disclaimer: I am an HP employee, but I don&#x
205.
by gjm11 1y ago
In context, the "at least 16%" is responding to someone who said 8%, and 16 just happens to be exactly twice 8. I suspect (though I don't know) that Yudkowsky would not have claimed to have a robust way to pick whether 16% or
206.
by gjm11 1y ago
I agree that it feels LLMish (though the LLMs learned it from humans and it's always possible that whoever wrote it just has that sort of style) but the dashes there are en-dashes rather than the longer em-dashes that LLMs seem par
207.
by gjm11 1y ago
Ohh, I see, you mean "mass" should have been "mass and energy" rather than e.g. that (mass,volume) should have been replaced by (mass,energy) or something. I confess I just ... take it for granted in this kind of context
208.
by gjm11 1y ago
Is that intended to be a correction? (I don't think the original statement needs correcting, other than by replacing "universe" with "observable universe" in both places.)
209.
by gjm11 1y ago
The event horizons.
210.
by gjm11 1y ago
How?

More ›