CarrotsHaveEars

AI this, AI that

I had a coworker a few months ago who “taught” me how to get more accurate output from Copilot by cloning into an “agent hub”, which is a repository of .agent.md files, and instruct it to load a bunch a “skills”, and apparently involves fiddling a bunch of options in VS Code settings.

Uh, at least tell us what kind of computer it is, or what the spec is like.

“Responsible use of AI”? Is that a joke? There is no such a thing as responsible use of AI. Responsible people would refuse to use AI.

But anyway, Linux is already AI. Sticking to any one of those BSDs is the only way out.

Pulseaudio.

Mr Torvalds and his now favourite AI toy.

What was really the problem? If you compile it the first time, the second time it only compiles the files you’ve changed. No matter how big your monorepo is, as long as it compiles overnight, there shouldn’t be any impact to the speed of work.

I wish that was true. However, look at the title, look at how people talk about this in general. “Yo, you heard Bun was rewritten in Rust? That is soooo lame.”

Really, the title shall just say “Bun vibe-rewrite”.

Please don’t mention Rust. Rust is a great programming language. The main culprit is they used AI to rewrite it. Rust does not deserve another rotten egg thrown at it because of this.

I think you mean doing that for more than once has no extra effect on an SSD. Shouldn’t we overwrite the unencrypted content for at least once?

Be careful posting articles like this. I’m quoting some parts of it before anyone jumps to conclusion without reading it.

What does the policy say?

The policy summarizes itself this way:

It’s fine to use LLMs to answer questions, analyze, distill, refine, check, suggest, review. But not to create.

Pre-arranged, non-critical, high-quality, well-tested, and well-reviewed code changes that are originally created by an LLM are allowed, with disclosure.

Moderation rules

You must disclose LLM-generated content. You can choose to not post LLM content, or you can choose to post it and disclose its origin. You may not hide LLM involvement.

And hard disks too. Fuck myself for thinking 4TB NAS drives were too expensive.

To be fair, I think a 32GB RAM should have costed $60 by now, if not for this AI shit.

Sorry, no idea about Podman but docker is fine.

It’s Arch Linux done right.

Believe me, OP, SBCs are much more expensive than secondhand laptops and the performance is crap. If form factor is in consideration, consider a mini PC in the shape of NUC, or just get a Mac Mini M4 (install Asahi). If you want something flat, you can buy a secondhand Macbook with the screen removed.

Always, always choose amd64 or Apple Silicon.

Don’t look for a solution that is technical where the problem is mental.

No worries. My bad.

miniserve, a popular command line program written in Rust to serve static files, also counts down (from three though) before launching the server, just like he counts down from five before someone rewrites your Hanoi Tower game in Rust.

http://crates.io/crates/miniserve

Why are you sharing your program with miniserve?

But the important part is that idiotic managers and shareholders buy it. Although they might have caused their own bankruptcy, people has lost their job, and the demand of labour is not even staying the same. It has caused unemployment after all.

Our company is like that and I don’t have the slightest problem with it.

Interesting. Let me try one.

Step 1, atomic bomb is very powerful
Step 2, everyone rushes to build atomic bombs
Step 3, ???
Step 4, everyone can use cheap nuclear power

The difference between dot-com boom and AI boom is that dot-com caused massive hiring at the beginning at ridiculous salaries, and dropped back to normal demand. It didn’t cause massive layoffs and force remaining workers to overwork.

Both managements are stupid. The dot-com management only lost their own money, but AI management made people leave their own country penniless in the worst case.