Ah I see, you’re saying it’s not “just a tool” because there is a whole political thing in companies about being forced to use it, which doesn’t normally apply to ordinary tools.

You may be right but it’s kind of irrelevant. When people say “it’s just a tool” they’re trying to assert that it isn’t going to fundamentally change anything about our jobs or society. It’s “just a tool” like a debugger or an IDE or a static analyser.

Your argument doesn’t really say anything about that assertion. So you’re sort of right, but in a way that nobody cares about.

And for what it’s worth the “it’s just a tool and isn’t going to fundamentally change anything” people are very clearly delusional (or they haven’t used Codex in the last 6 months). There is absolutely no way this is “just a tool”. Most people I speak to really have stopped writing code by hand. It’s going to completely screw young people looking to get white collar jobs. A debugger never did that.

Yeah he has rediscovered long polling.

This is the worst argument against “just a tool” I’ve heard.

Several of those are fairly borderline. You would think Drew Devault of all people would understand that its ok to not be 100% squeaky clean.

I’m not saying that list doesn’t include a lot of creeps, but maybe he shouldn’t have included “donated to an organisation that includes someone who is antagonistic” and “is an average American Christian” along with the “fervent defender of paedophilia”.

The fact that CSS only supports multiline comments doesn’t mean multiline comments are good, it means CSS is badly designed.

Close notepad.exe. Open a real editor. Select the text you want to comment out. Press Ctrl-/

Yeah but… Don’t. Multiline comments are a bit of an anti-pattern at the best of times. This is a war crime.

AI is not capable of novel or complex stuff,

This seems to be both untrue and mostly irrelevant.

The recent FLT proof is clearly complex and novel.

But also a huge amount of programming is only barely complex and not really novel at all.

The biggest issues I’ve seen with AI so far are:

• The slop comments are just painful to read. I think this might mostly be Claude though, I recently switched to Codex and it’s way less annoying.
• It is so benchmaxxed to “succeed at all costs” it will often do stupid things just to make the test pass or whatever. To be fair a lot of humans are like that - “I dunno but it seems to work” attitude.

I expect they are for GitHub. GitHub is really well known and huge so it makes sense to set up specific scrapers for it.

This seems reasonable. Cry all you want but if even smaller sites are getting overwhelmed with bots imagine what it’s like for GitHub.

What are you talking about?

Ah the classic “it’s just maths therefore it can’t really think” nonsense. You’re just dirty water so you definitely can’t think.

Python is a poor choice IMO. It’s just too slow, even for web servers. Everyone says “oh it’s IO bound or database bound; Python doesn’t matter”) but in my experience it does.

Better choices:

• Go - very easy. Build in web stuff. Fast. Trivial to deploy.
• Typescript, especially via something like Deno’s Fresh. Much faster than Python, better type checking, and best of all you can use TSX-based templates instead of shitty text-based ones like Jinja.
• Rust. Definitely the hardest option, but on the plus side you will learn Rust and that’s a big plus.

They can be guided. For example I asked Claude to benchmark some SystemVerilog.

First it just did time ./run.sh and I was like “no, idiot. That’s going to include the compilation/startup overhead.”

So then it said “I’ll just increase the simulation time to amortise the startup cost.” Again I told it that was bad and to actually measure the time in SV.

So since SV has no wall-time functions, it used $system("date") or some bollocks like that. No! I said - use DPI-C.

Finally it did the right thing. Would it have been quicker to do it myself? Absolutely not.

Current AI is at lazy junior engineer level (but sped up 10000 times). You just have to make sure they don’t make lazy dumb decisions. E.g. they’re going to do everything in Bash if give half a chance.

Tbf to AI I think they’ve learned a lot of this from humans. I bet if you searched for $system in SV you will find a lot of hand crafted horrors. I’ve seen human-written C code that found an ELF symbol by system("nm ... | grep ... | awk ..."). I shit you not.

Yeah… when the only documentation is CLAUDE.md…

Still you can see why right? Improving docs/understandability for AI is good for them - they get more capable AI. Improving it for humans only helps other people, which is less of a priority.

Really? Because some of his hand-picked bug examples are “syntax highlighting is sub-par on GitHub.com”, “Google Analytics’ map view shows some cities in the wrong place”, “a weapon in a game works differently than intended in a way that isn’t noticeable to players”, “a Google hangouts call didn’t work once”.

Technically bugs but… Yeah let’s see his perfect software.

Tenure doesn’t mean much.

Well in a sense that is true. This guy is pretending like he’s the only one that sees bugs. In reality he’s the only one that isn’t able to understand priorities. If you’re constantly reporting every bug no matter how trivial, that is a problem. You’re wasting people’s time.

Never is a strong word. In cases where you don’t need strong cryptographic or statistical properties it’s totally fine. That’s 99% of cases.

Nobody is modding a single byte. You’re going to have a 32-bit PRNG at least. 32-bits mod 100 is easily random enough for almost every use.

I guess it’s a good thing to be aware of though.

Every core, honestly

Oh hi Claude.

Yeah these types of posts are kind of useless because you can never tell if it’s just because they’re giving off red flags.

What feedback did he get from interviews? Do people know you can ask for feedback?