qsort

20 ms·

241.
by qsort 1y ago
Thanks for sharing, that was a good thread!
242.
by qsort 1y ago
You sample only from tokens that could possibly result in a valid production for the grammar. It's an inference-only thing.
243.
by qsort 1y ago
Isn't that pretty much in line with what people were expecting? Is it surprising?
244.
by qsort 1y ago
SWE-Bench is also not stellar. "It's important to remember" that: - they are only evals - this is mostly positioned as a general consumer product, they might have better stuff for us nerds in hand.
245.
by qsort 1y ago
I'm the first to call out ridiculous behavior by AI companies but short of something massively below expectations this can't be bad for openai. GPT-5 is going to be positioned as a product for the general public first and foremost
246.
by qsort 1y ago
Tangential, but for some reason P vs. NP attracts an ungodly amount of cranks, probably because as far as open problems of that importance go it's easy to understand the question.
247.
by qsort 1y ago
Buffalo mozzarella is actually a trap on pizza. It contains more water and it tends to dampen the dough as it melts. Fiordilatte or Treccia are actually better, even though they're cheaper.
248.
by qsort 1y ago
All three major labs released something within hours of each other. This anime arc is insane.
249.
by qsort 1y ago
pelican when
250.
by qsort 1y ago
It's extremely hard to measure productivity correctly and self-reports are worthless. I don't think AI tools are a net negative in the average case (people are definitely indexing too much on that goddamn METR article) but &quot
251.
by qsort 1y ago
People are using different definitions of "vibe coding". If you expect to just prompt without even looking at the code and being involved in the process the result will be crap. This doesn't preclude the usefulness of models
252.
by qsort 1y ago
We could say it's all you need...
253.
by qsort 1y ago
Vim is very likely to be an editor that complements a full IDE if you're just doing something quick. If you asked me I'd also check Vim even though I'm usually on Jetbrains.
254.
by qsort 1y ago
Probably not as popular, but C# is absolutely up there, even though it doesn't get a lot of mindshare. I work for a small firm that does consulting for enterprise customers and I see a lot of C# around. Europe especially is basically M
255.
by qsort 1y ago
> The top 10% is even better with LLMs, they know much more and they are more productive. But the rest have just resulted to turning in clear slop with no care. This is interesting, I'm noticing something similar (even taking LLMs o
256.
Tao on “blue team” vs. “red team” LLMs(mathstodon.xyz)
542 points by qsort 1y ago|171 comments
257.
by qsort 1y ago
I think I partially agree, your first paragraph is exactly how I feel. Boilerplate and trivial stuff absolutely should be automated. It's also true that people have been pushing a narrative where programming is some dark art and you sh
258.
by qsort 1y ago
Appreciate the response, sorry if I'm being a bit direct but as you probably know that's the style that works the best on a forum like this one. I'll keep tabs on you guys, my DS colleagues might be interested in the project.
259.
by qsort 1y ago
You're welcome to suggest one that has 10% of the functionality that with PyCharm you take for granted. Graphical table creator? View and export ER schemas? SQL Syntax and autocomplete in .sql files AND within literal strings in your c
260.
by qsort 1y ago
I don't want to dunk too hard on this as it seems to be reasonably well made, but how are we making a data science IDE without a good SQL client? I might be biased but that's a major part of the workflow. You're already los
261.
by qsort 1y ago
I don't think AI lives up to the current hype but this article is garbage. They're obviously talking about the METR paper, but the main takeaway according to the authors themselves was that self-reporting productivity increases is
262.
by qsort 1y ago
I'm also seeing benefits, and I'm on pretty much the same stack as you. The METR paper I'm citing was misquoted a lot (N was small, and it wasn't a favorable setup for AI tools), but the most important finding was that i
263.
by qsort 1y ago
And then you have research that says they're both full of shit. The article is perhaps a bit shallow, but it's spiritually correct: there's a lot of uncertainty and people who claim they figured it all out are mostly spewin
264.
by qsort 1y ago
It's also 50 bytes longer than the original. More LOC only because my Vim formats on save. Whenever somebody comes up with some big brain idea with macros, ORMs, DSLs, 180 IQ templates, language extensions that even Haskell nerds would
265.
by qsort 1y ago
Is it really that much better than this: #include <X11/Xlib.h> #include <stdlib.h> int GetKeyCode(Display* d, char* s) { return XKeysymToKeycode(d, XStringToKeysym(s)); } int main() { Displ
266.
by qsort 1y ago
Artificial intelligence is an umbrella term. Serious people already use the name of the actual technology. LLMs, support vector machines, XGBoost, linear regression, expert systems, prolog interpreters... Just say what you're actually
267.
by qsort 1y ago
This is interesting but I'm not sure I'd want it as a default behavior. Managing the context is the main way you keep those tools from going postal on the codebase, I don't think nondeterministically adding more crap to the c
268.
We do stand at the edge of a pronounced crisis of identity(twitter.com)
3 points by qsort 1y ago|0 comments
269.
by qsort 1y ago
I'm obviously not qualified to judge the artistic value, but one of the picture has a diagram where an arrow labeled "Perestroika" connects a box labeled "Collapse of the Soviet Union" to the word "Corporate me
270.
by qsort 1y ago
It's still tabula rasa -- you're just initializing the context slightly differently every time. The problem is the constant anthropomorphization of these models, the insistence they're "minds" even though they aren&

More ›