59nadir

17 ms·

121.
by 59nadir 6mo ago
I agree entirely with your statement that structure makes things easier for both LLMs and humans, but I'd gently push back on the mutation. Exactly as mutation is fine for humans it also seems to be fine for LLMs in that structured mut
122.
by 59nadir 6mo ago
> he'll likely go more and more towards vibecoding again I think "more and more" is doing some very heavy lifting here. On the surface it reads like "a lot" to many people, I think, which is why this is hard to r
123.
by 59nadir 7mo ago
Personally I read the GP post with more emphasis on this bit: > What is (not) here to stay are the techbros who think every problem can be solved with LLMs. LLMs are in all likelyhood here to stay, but the scumbags doing business around
124.
by 59nadir 7mo ago
That's weird, because every time I see someone even talking positively about Claude Code they always seem to mention they're hitting their 5 hour limits in 2-3 hours all the time, they're hitting their overall limits all the
125.
by 59nadir 7mo ago
> Sacrifice civilian infrastructure as the only viable bombing target. I'm imagining the air crew going "Huh, there are no clear actual targets to bomb. Hey, Cleetus, command won't be happy about us not bombing anything a
126.
by 59nadir 7mo ago
You do know they have an agent that runs in the terminal, right?
127.
by 59nadir 7mo ago
Heh, fair enough. To me this comes off as "I'm unable to write it myself [possibly because I've outsourced my thinking too much]" , to be honest, but I'm not going to argue; you're the one who presumably wants
128.
by 59nadir 7mo ago
Eh, fair enough. 1.5k is reasonable. Have you tried just writing it yourself instead? It's likely it'll be less than 1k lines and you should have no problems writing an implementation yourself if you understand the structure of th
129.
by 59nadir 7mo ago
> How about Claude Code? 100% of it was vibe-coded according to its creator. Agents are trivial to make. I don't know whether that means they're not "serious", but it's exactly the type of thing you can make your
130.
by 59nadir 7mo ago
"Well structured + tested". Who would know? The diff is almost 200k changed lines. Good on them for saying no to this nonsense. There's a good chance the actual needed implementation is less than 20k lines (I've found th
131.
by 59nadir 7mo ago
Given how Windows 11 is going I would guess "You know what's a lot harder? Not trending downwards." would be the more obvious one.
132.
by 59nadir 7mo ago
Intrinsically simple and straight forward problems are easier to secure even with mediocre or bad code. They've already shown that Opus 4.6 can find and report on very sophisticated security issues[0] so I'm not sure that analysis
133.
by 59nadir 7mo ago
I deleted my account a few years ago, I might actually create one now, it'll be preferable to HN if they stick with this new rule.
134.
by 59nadir 7mo ago
I think one thing that goes unmentioned is that maybe code quality is really not that important for trivial things, because they can be trivially reproduced if need be. I would argue Claude Code is exactly such a project; coding agents are
135.
by 59nadir 7mo ago
All of the above at exactly the token cost that it requires for you . Anything general is always going to be worse for specific use cases, and agents from these big providers are very general. They'll spend tons of tokens doing things
136.
by 59nadir 7mo ago
I think it's good that it's out there, and I wonder why Anthropic have been keeping it closed source; clearly they can't possibly think that the CC source code is a competitive advantage...? Agents in general are easy to make
137.
by 59nadir 7mo ago
No, I meant that it is other people's code run through a tumbler, as is all LLM-generated code.
138.
by 59nadir 7mo ago
> I don't like reading leaked code Don't worry about that, the code in that repository isn't Anthropic's to begin with.
139.
by 59nadir 7mo ago
Closed source AI is mostly US with all of the nonsense that entails, on top of being closed source. I don't think this is the even comparison you seem to think it is. The rest of the world probably doesn't look at this equation an
140.
by 59nadir 7mo ago
Personally it would make the choice to say no to the entire thing a whole lot easier if they self-reported on themselves automatically and with no recourse to hide the fact that they've used LLMs. I want to see it for dependencies (I a
141.
by 59nadir 7mo ago
This reads like a "You wouldn't download a car!" ad but it's trying to scare you into using AI instead.
142.
by 59nadir 7mo ago
Vibe copy is a hell of a drug.
143.
by 59nadir 7mo ago
If it was any good at sales I'm pretty sure a company I did a contract for would be thriving by now. Instead they have a product that is ~500 times faster than the competition, with better UX for the most common activity in that field
144.
by 59nadir 7mo ago
We've seen public examples of where LLMs literally disable or remove tests in order to pass. I'm not sure having tests and asking LLMs to not merge things before passing them being "easy" matters much when the failure mo
145.
by 59nadir 7mo ago
> I've been looking at it for many years - they are simply unable to leave the dream era. Sit down and do the work needed to get Ruby running properly on Haiku instead of sitting here complaining and basically admitting that you&#x2
146.
by 59nadir 7mo ago
That's a fair observation and one that I don't really have an answer for. I can say from personal experience that I believe that shipping nonsense code has never been faster. That's just an anecdote, obviously. We need a bigg
147.
by 59nadir 7mo ago
Software developers have spent decades at this point discounting and ignoring almost all objective metrics for software quality and the industry as a whole has developed a general disregard for any metric that isn't time-to-ship (and e
148.
by 59nadir 8mo ago
I don't think a lot of people are really worried that LLMs will successfully replace them, but they might still get let go because the people in charge think they can replace people with LLMs. These two scenarios don't imply t
149.
by 59nadir 9mo ago
Thank you for actually posting an example that people can look at; I think most other responders misunderstood the post as asking for more pointless anecdotes filled with superlatives and "trust me bro" sentiments.
150.
by 59nadir 9mo ago
My bigger point was that not everyone who is skeptical about supposed productivity gains and their veracity is in competition with you. I think any inference you made beyond that is a mistake on your part. (I did do web development and dist

More ›