C’mon Europe! Distill Anthropic and OpenAI like the Chinese models and then open source your model, weights and training data!

I wouldn’t say the industry as a whole is toxic to profit. Nvidia, for one, has been almost literally minting money. And then there’s AMD. And, uh…

Right, but this uterus isn’t in a woman.

I think the incels want the sex part and not necessarily the kids part tho – this would give you the opposite

Brave New World, here we come…

I think big tech has proven that it cannot be trusted.

I’d go a step further and say that if big tech must control the internet it would be better to not have an internet at all (and that’s coming from someone who remembers life without it well and currently uses it a lot). But given the current situation I would hand back any true productivity gains, casual enjoyment, etc to take away the global narrative control that a tiny handful of companies have.

Unfortunately it seems to be the “expert” thing that’s important, even more than the “bird watching” thing. As it turns out, training your brain for hundreds or thousands of hours over the course of years is good for it.

Thank you. Something stirred in the back of my mind when I read the title. Would have jumped down the rabbit hole if not for this comment 🙂

They should have vibe-coded their accounting.

I think we’ll soon see exactly how good “reasoning” models actually are, as the next generation of problem-havers uncover things that were never discussed on SO or elsewhere in-depth.

Definitely my favorite new community in a while. Keep it up!

I’d rather see better discovery tools and better community/account migration tools. Id be worried about topic-specific instances potentially backfiring by concentrating too much influence for a given set of subjects on the “preferred” instances

Also one of the first examples of a long running, stable (mostly, probably) egalitarian civilization.

I used to think Peter Thiel was a Nazi too.

I mean, I still think he is, but I also used to.

RSS still leaves data in silos. If I love the content on site.com and they decide to put it behind a paywall, enshittify it, or the site goes down, that content is lost. On the fediverse, as long as it has propagated across the network it can be found on other instances (the content being comments in this case, not offsite-linked materials).

Maybe a browser plugin/extension would be better in your case then? That would actually probably be not too hard to vibe-code if you’re so inclined.

I’ve seen this ask come up a number of times now – and to a certain extent, the Piefed and Mbin frontends do kinda to it already in that you can have your microblog and threaded content in the same UI. But before taking that further with pixelfed/loops integration as well, ask whether it’s really a good idea. There’s a reason why Twitter, Facebook, and Reddit are separate sites, and why none have tried to implement the UI or feature set of the other. I would make the case that the fediverse would really benefit from a portable user IDs auth system that could be used between different services or instances (which Nostr kinda does, and there are other zero-knowledge proof-based services that could be integrated), but with regard to content display and just interacting with content, having 3 different tabs open with the 3 different kinds of content/conversations just makes a lot more sense to me.

I wonder, do instance admins or community mods ever have (virtual) meetups? That might be therapeutic and could possibly expose problems before they become too big to solve.

At least we can all agree that Hawaii is awesome

Yeah, one real concern is that bottom-of-the-barrel lawyers will continue to just use their $20/month chatGPT subscription, and not something more lawyer-centric that will (eventually) be able to weed out the true stupidity almost 100% of the time.