Sometimes it really feels like you need to do extensive background checks on everyone involved and be an expert at politics and sociology and biology and geology in order to not accidentally choose an unethical software.
@baralheia Don't you have uncertainties with a lot of other factors, too? How can you know a developer didn't just copy-paste code from StackOverflow? Or that they spent a lot of thought on solving the issue instead of going with the first solution that crossed their mind? Or that their ability to program wasn't hampered by something bad happening in their life? What if GitHub kicked them off the platform all of a sudden? What if there's an oversight like github.com/ValveSoftware/steam-for-linux/issues/3671 ? There's always uncertainty.
My general approach to using software and platforms is to think about the worst realistic scenario that could happen and how badly I would be affected so I can be prepared for it. I was talking about Zed today and people raised concerns because the developers probably use AI. Right now, it's a significant improvement over VSCode. The worst realistic scenario would be a drop in quality so bad I'd have to stop using the editor again. Maybe some destroyed files if we want to be really pessimistic. How bad would that be for me? Actually not much. My projects are backed up with git, there are backups of my system, and I can switch editors at any time. So I don't care a single bit if they use AI until there are actual manifestations of its bad effects.
Let me explain why "they use AI" is more often than not a meaningless statement to me. It certainly can be useful to point out — but only if it's truthful. The problem is that it very often isn't. People tend to exaggerate the degree AI is involved because they despise it, and so the dev writing "I used AI to bootstrap this project" on their blog becomes "this software is slop written by AI". There are many ways AI might be part of a project. Maybe the devs only used it to generate themselves a more tailored boilerplate. Maybe they do use it more, but hold themselves to high standards to ensure nothing bad happens. Maybe they vibecoded the whole thing. Maybe they accept AI-assisted contributions, but don't use it themselves. There's a whole lot of nuance that gets condensed into "absolutely zero AI involved" and "irredeemable trash". If you don't care about it, then so be it, but I do.
It's finally my turn to share a screenshot where Google's AI overview reacts to the query as a prompt instead of summarizing the results. I only wanted to check if it was "give" or "keep me company".
@petafloppa Fash-leaning techbros also use computers. Does this mean we're automatically fash-leaning techbros? I'd rather go by evidence of direct endorsement of fascism, not this.
I switched to oxfmt and oxlint and those run perfectly. They even have features I was missing in Biome, so it was a worthwhile change even if I end up moving away from Zed again
People say "this software is/might be developed with AI" as if that magically means any actual experiences with the software don't matter. If it runs well, I will use it. If it doesn't — like anything Microsoft produces — I won't. It's that simple.
@antimu0n@catraxx@zaire I see all sorts of issues with AI as a software developer. But as a user, I don't see why I would care what people use to develop their project. All that matters is if the software runs well. Windows 11 is constantly getting botched updates now, so that's bad. Right now, my impression of Zed is a good one and it's running way better than VSCode, so I'd be stupid to stick with VSCode just because of suspicions that aren't reflected in the experience I'm having.
Volpeon