Keep in mind. I'm as about as skeptical of this so-called AI as anyone else. Even though I have self-hosted them at home. I don't generally run them. Because they are largely useless to me.

Still though, we see people having for lack of better words. Very strong, lower knowledge reactions to the discussion. If you feel that's pretentious, then honestly what terms would you suggest for those well beyond healthy skepticism?

Why is the existence of this project an issue because AI was used, it works and works well for me better than any other rss reader I've used. It wouldn't exist otherwise and I would be having a worse experience. What it made in 2 days is better than I could've hoped to achieve in months and my goal isn't to be a dev, it's to use the app I have in my head. To have the convenience I am seeking, which is why I seek out opensource apps in the first place. It's never stable, always buggy, long before AI, but it fulfils my need good enough for free...

The idea that because you have self-hosted before it makes you somehow qualified to speak about this is just you sitting on a high horse and throwing anecdotes at me condescendingly, it means nothing.

We can debate about the level of qualification all day long. But I'll tell you one thing, it makes me more qualified than someone who hasn't. Which was the point you're trying to deflect from. And talk about condescending. Either way I'm not here to troll or caused drama. I gave my two cents and you can get back to your Purity or whatever I'm out. Can't have a reasonable discussion with someone who behaves like this.

There is no credible evidence self-hosted or corporate AI delivers on any of its promises, it is just a woodchipper to obliterate our humanity in so we can spit out bloody chunks and point at them and say "wow, almost like the real thing!".

AI is a snake eating itself/a massive theft of the commons in an attempt to privatize it, and once the public's shared wealth of open information has been strip mined and then drowned in AI submitted crap to push things that direction, all of this collapses into complete and utter junk.

If we decide to actually go through with this dumb shit and get to that point I will never stop raising the point that this was all a self-defeating choice we made not any sort of "inevitable" new powerful technology we all had to bow down to and become dependent on.

Specifically I was taking with regard to self hosting. I have self hosted before. Have you? Since you're acting like a subject matter expert who's sure of themselves. Can you explain why that even with having no hoses or anything run to my desktop, that it worked just fine? Yes it's inefficient. So were computers at one point. Filling buildings and consuming kilowatts of power. They got better.

Frankly I really don't like it being called ai colloquially either. I don't think what is being paraded around us truly conscious etc. But the reactionary behavior is what leads to a lot of shit shows. Purposeful misunderstandings etc. Like I said, you probably mostly agree with me. Whether or not you understand that.

Whatever we think of ai. It doesn’t matter. It’s here. It’s not going away.

Your entire argument hinges on a delusional, near religious belief in the inevitability of AI.

It is a broken technology that uses a massive amount of energy to serve up results that are often catastrophically wrong.

You cannot paint the future with such decisiveness without evidence, and this is precisely what is leading to a collapse of AI valuations. It is all just vibes and no substance.

Whatever we think of ai. It doesn't matter. It's here. It's not going away. And trying to reactionarily be pure and untainted is only going to cause problems for ourselves and everyone else.

I think the one thing that could be universally agreed upon at least in the fediverse. Is that people shouldn't be using proprietary models run in remote data centers by fascist oligarchs.

Sadly in those places we cannot have a nuance debate about what we should do in response. My issue with Torvalds and others isn't specifically that they are using ai. I think someone with his background and experience, can turn out good code faster using AI. My problem is that he's likely currently using a big proprietary model to do that.

We should all be using local models if any. Which is something that is starting to happen. Even today IBM was talking about it. Sure they were talking about it in reference to them selling the hardware for it to their customers. But still self-hosted nonetheless. And on top of that ethically trained models. I think all of this will come in time. Even with all the toxic reactionary virtue signaling.