- Am I stupid if I don't get it? What is the intended end state? What does "ungroup operands" mean?
- One of the mental frameworks that convinced me is how much of a "free action" it is. Have the LLM (or the agent) churn on some problem and do something else. Come back and review the result. If you had to put significant effort into each query, I agree it wouldn't be worth it, but you can just type something into the textbox and wait.
- I got nerd sniped and tried to reconstruct the games. The notation is weird and they aren't using modern conventions (a1 is dark, queens on the d-file, kings on the e-file, white goes first.)
Also, as the article mentions there are a few errors. With a bit of deduction this is my best attempt at reconstructing the first one:
In keeping with the theme of exciting new technology, I tried giving the problem to Opus 4.5 but it seems to hallucinate badly: https://claude.ai/share/299fb10e-8465-41b3-bad5-85500291ed67
- I care very little about fashion, whether in clothes or in computers. I've always liked Anthropic products a bit more but Codex is excellent, if that's your jam more power to you.
- I believe the argument isn't that ancient statues were ugly, but rather that reconstructions are ugly (unfortunately this has been used to argue against the now ascertained fact that ancient statues were indeed painted). Purely subjective judgement from someone not trained in the arts: that photo of the Augusto di Prima Porta doesn't look like a great paint-job. The idea that, like the statue itself, the painting must instead have been a great work of art lost to time seems solid to me.
- This is very interesting if used judiciously, I can see many use cases where I'd want interfaces to be drawn dynamically (e.g. charts for business intelligence.)
What scares me is that even without arbitrary code generation, there's the potential for hallucinations and prompt injection to hit hard if a solution like this isn't sandboxed properly. An automatically generated "confirm purchase" button like in the shown example is... probably something I'd not make entirely unsupervised just yet.
- I agree saying "they don't think" and leaving it at that isn't particularly useful or insightful, it's like saying "submarines don't swim" and refusing to elaborate further. It can be useful if you extend it to "they don't think like you do". Concepts like finite context windows, or the fact that the model is "frozen" and stateless, or the idea that you can transfer conversations between models are trivial if you know a bit about how LLMs work, but extremely baffling otherwise.
- I think it's more of a thread-bound dynamic rather than HN as a whole. If the thread starts positive you get "AGI tomorrow", if the thread starts negative you get "stochastic parrot".
But I see what you mean, there have been at least a few insane comment sections for sure.
- It's one of the failure modes of online forums. Everyone piles on and you get an unrealistic opinion sample. I'm not exactly trying to shove AI into everything, I'm weary of over hyping and mostly conservative in my technology choices. Still, I get a lot out of LLMs and agents for coding tasks.
- Sure, that's a great point. If the LLM code/ideas they come with are actually valuable, they tend to fall into the first bucket though.
I'm not saying we need to dismiss people for using LLMs at all, for better or for worse we live in a world where LLMs are here to stay. The annoying people would have found a way to be annoying even without AI, I'm sure.
- As someone who does consulting, it's more about the attitude than the tool itself. Clients trying to understand the problem by themselves with whatever tools they can use are generally well-disposed and easy to work with. Those who email you stuff like "Why don't you have chatgpt do this???" as if it's a revolutionary thought are mostly a PITA. I assume doctors feel largely the same.
- Everyone is doing this extreme pearl clutching around the specific wording. Yeah, it's not 100% accurate for many reasons, but the broader point was about employment effects, it doesn't need to completely replace every single developer to have a sizable impact. Sure, it's not there yet and it's not particularly close, but can you be certain that it will never be there?
Error bars, folks, use them.
- Yeah, I think this is one of the (few, rare) cases where the "official" academic way of teaching the subject is actually baggage and not really aligned with what's practically useful.
Compiler courses are structured like that because parsing really was the most important part, but I'd say in the "modern" world once you have a clear idea of how parsing actually works, it's more important to understand how compilers implement language features.
Even if you want to implement a compiler yourself, "Claude, please generate a recursive descent parser for this grammar" is close to working one-shot.
- I broadly agree with you but the AI demand part is more nuanced. Demand is very large for consumer-ish apps like ChatGPT and Sora, but even Claude, which is huge for coding and arguably the best model right now, is tiny. I've had an interaction with someone nontechnical who, after seeing the Claude logo bounce, asked me why my ChatGPT had fireworks.
Yeah, the demand is there, but I have a hard time believing nontechnical people are clamoring for Copilot, they likely don't even know such a thing exists. The market is insane right now.
- Do you frequently use Python for competitive programming puzzles? I've done it a bit in the past, and everyone always used C++.
- > I always code with the mindset “the compiler is smarter than me.”
Like with people in general, it depends on what compiler/interpreter we're talking about, I'll freely grant that clang is smarter than me, but CPython for sure isn't. :)
More generally, canonicalization goes very far, but no farther than language semantics allows. Not even the notorious "sufficiently smart compiler" with infinite time can figure out what you don't tell it.
- Anthropic? The AI people?
- > Complexity of traveling salesman problem depends on what you consider as the problem space - number of cities or number of connections between cities.
lmao what?
- I love Dijkstra's writing, but I don't think this is his strongest piece. In general parlance, when we say "by piegonhole" we mean "any variant of it". I'd still call what he's doing "piegonhole" lol. You can even further generalize it, e.g. by making expected value arguments.
This is not uncommon: we can say that "by the fundamental theorem of algebra" two polynomials of degree N that agree on N+1 points are identically equal. "By induction" includes Cauchy induction, sometimes with "this and that are the same" we mean "up to isomorphism" and so on.
The advice he ends on is extremely solid, though:
The math will always math.The moral of the story is that we are much better off with the neutral, general standard procedure: name the unknown(s) so that you can formalize all conditions given or to be established and simplify by formula manipulation.
((a & b) $ c) @ d
If it's reduced below both & and @ then it becomes:
(a & b) $ (c @ d)
I think conceptualizing parentheses as "increase priority" is fundamentally not the correct abstraction, it's school brain in a way. They are a way to specify an arbitrary tree of expressions, and in that sense they're complete.