zahlman

33 ms·

931.
by zahlman 5mo ago
> The politics were mostly fine, but the students, especially those post-COVID, were the problem. I'm not sure this distinction can be made, really. > And don't get me started on the false disability claims (see [0] for a ta
932.
by zahlman 5mo ago
I don't feel like trying to get information from an LLM is a different kind of "communication" from, say, writing code. And I don't use INTERCAL, so.
933.
by zahlman 5mo ago
>I'm not even sure most english is spoken by native speakers anymore Per Wikipedia, only about a quarter of English speakers have it as their first language.
934.
by zahlman 5mo ago
I assume "require" is meant in the terms-of-service sense.
935.
by zahlman 5mo ago
I've thought about making explainer videos for math and CS concepts (with animated diagrams etc., you know). But I really don't want my voice and image out there any more than necessary / they already are. Now I'm wonder
936.
by zahlman 5mo ago
Indeed. I've heard a few compositions that I knew were AI-generated and still thought were pretty good.
937.
by zahlman 5mo ago
> While we still require creators to manually disclose when they use realistic AI, we want to make the process more seamless and reliable. Starting in May 2026, we’re rolling out new internal signals to help identify AI-generated content
938.
by zahlman 5mo ago
I feel like, if an LLM can one-shot a say 500 LOC project from a single-line prompt, and doesn't require the user to make any choices, and the result is actually acceptable, then 499 of those LOC are definitionally "boilerplate&
939.
by zahlman 5mo ago
> A lot of people think a lot of things, but I don’t think the majority of people think the point of using LLMs is so they can produce low-quality code. Hence "seem". Of course people are not in the habit of describing their pr
940.
by zahlman 5mo ago
> Clean Code (the book, not the concept), etc. Thank you for drawing the distinction. I've noticed that while ~80% of Uncle Bob's ideas are good, he doesn't seem to have been very good at implementing them, and he often
941.
by zahlman 5mo ago
>the code pours out sometimes faster than I can write. Meaning that you type the code faster than you would normally type prose? Or just what?
942.
by zahlman 5mo ago
> but I fear that workflow is just chat, search, and being a rubber duck for my thoughts. That's still a lot of benefit , though. I have to agree with Patrick McKenzie on this one ( https://x.com/patio11/status&
943.
by zahlman 5mo ago
Ah, Trey Hunner's work. Definitely among my inspirations. I can't find it now, but I seem to recall he had a really good explanation on Stack Overflow of how to read nested list comprehensions that definitely influenced the way I
944.
by zahlman 5mo ago
For reference, the general form (so that I can use letters to refer to parties and avoid convoluted phrasing): (Axiom: B has done X.) A (to B): You have done X, and you should not have done so. C: I note here that certain prior actions of A
945.
by zahlman 5mo ago
The point is that nothing you are saying here has anything to do with Python or its design and would be equally true of any programming language with a comparable ecosystem.
946.
by zahlman 6mo ago
Compilers not doing something is not a demonstration that they are not actually allowed to do that thing.
947.
by zahlman 6mo ago
> illegal instructions... were often used by programmers Intentionally, with an expected effect? I'd need a citation for that.
948.
by zahlman 6mo ago
> Here's a way weirder example: Well, yes; but when the C standard authors wrote like this, they surely had in mind "the reads could be in either order, therefore the output could display the polled values in either order"
949.
by zahlman 6mo ago
I understand the instinct, but that's a bit too prescriptive for me. https://en.wiktionary.org/wiki/learnings
950.
by zahlman 6mo ago
Has Rust code generally been found shorter than C++ in practice? I don't see an obvious reason for it.
951.
by zahlman 6mo ago
You might even say it finds a way.
952.
by zahlman 6mo ago
My point was the "with hashes" part. You aren't in fact "trusting" NPM to ensure that old versions aren't replaced if the package installer is verifying the hash.
953.
by zahlman 6mo ago
> the models will be good enough in social engineering Never mind questions of how good the models will/can get. I'm confused why people expect that, in principle , models getting really good at social engineering would have s
954.
by zahlman 6mo ago
> and trust that NPM at least gets their "you can't republish to an old version" bare-minimum constraint right? ... Does NPM not create full lockfiles, with hashes and pinned transitive dependencies and everything?
955.
by zahlman 6mo ago
> the good old days when people would hand out low privilege Linux accounts and rely on the kernel to prevent privilige escalation. ... Did that actually stop? ... Is that not the purpose of being able to make multiple accounts (even o
956.
by zahlman 6mo ago
... Deno and Bun manage packages? I thought they were just runtimes.
957.
by zahlman 6mo ago
In fact, attacks like Shai-Hulud explicitly attempt to get into PyPI, and have succeeded to a lesser extent. But aside from the package-size / -complexity issue pointed out in a sibling comment, PyPI also tries a fair bit to monitor fo
958.
by zahlman 6mo ago
> Just one line in AGENTS.md and CLAUDE.md: "Start by reading PROJECT.md" . > And PROJECT.md contains... …Why not just use that PROJECT.md as the AGENTS/CLAUDE.md?
959.
by zahlman 6mo ago
Do keep in mind, the user has to be someone who can conceive of the idea of having a script instead of just having the job done, and also needs to have some semblance of understanding how to run it.
960.
by zahlman 6mo ago
... If it's been like that for 10 years then it can hardly be blamed on "AI".

More ›