Post

Can't code? I'd say ask a friend, but at the rate AI is improving, maybe you won't even need to. Someone built a functional Appview with replit overnight. bsky.app/profile/doll...

Doll@dollspace.gay

The Appview code should have everything blueskys appview does, the only problem it is running into is getting the right numbers for database connections and pooling. github.com/dollspace-ga...

17 reposts195 likes

man you are just speedrunning a marathon of dumbass decision making. defending fascists, drowning out trans voices, vocalizing any support of the plaigarism environment destructor machine. there aint much left to dig in that shit garden.

If you're architects instead of moderators, maybe fix the Only Posts feed that's been repeatedly showing the same posts from 4 days ago since last week? We've logged out, logged back in, cleared caches, refreshed. Still, there they are. Thanks.

Nah. I'd rather not promote technologies that promise to lobotomize anyone who uses them. If you're just going to go "fuck learning, thinking and doing for yourself, just have Big Tech do everything for you", then I'm not sure why you're even bothering with this whole federated social media thing.

...you know what, I think I just figured out why Bluesky has been alternating between inflammatory comments and cute kitties/food porn as of late. You've delegated all official bluesky communications to "Artificial Intelligence" trained on twitter, haven't you? Am I right?

I definitely don’t worship it, but I am a pragmatist about learning how to use technologies that can’t be uninvented before they use us.

When you're watching a machine at your job running unsupervised and presenting a hazard, you're usually expected to take the pragmatic, responsible action and turn that thing off.

wait so asking as someone who has written neural nets to learn how they work how in the name of fuck do you think a glorified equation is going to "use people" because AI is just that, an equation that maps values to expected outputs, it doesn't think, feel, or have goals

this doesn't mean anything you desperately need to log off, they are just Expensive Autocomplete they are not alive I am afraid that you are cooking your brain with tech industry bullshit, this isn't healthy

Also AI isn't improving, LLMs are a dead end technology, the hallucination problem cannot be overcome, and the only way to increase the sophistication of the models will be AI coprophagia where they ingest their own refuse which will rapidly lead to the end of whatever miniscule value they have.

ah yes everyone should get on board with the technology that generates shit code. there's no uninventing the shit code generator. might as well not get left behind and generate shit code with the shittiest of them.

Lmao the resource guzzling 'get things wrong' machine isn't something that warrants pragmatism. It exists in an attempt to solve the problem of paying people to do labor. And it's never going to do that. It's a huge bubble of waste. You really are Elon lite.

AI isnt going to use us. It cant. It litterally cant. It's a modal machine. It just spits out the most common answer. Maybe just dont use technology that wastes resources and damages the enviroment and cant even give a correct answer

You don't get to go "oh but I only support the good stuff, not that" on this. Anyone who has chosen to be a booster for this tech has chosen to be a booster for this too, as it's integral to how it works

we've gone from the rather dumb "techbros insisting everyone should learn to code" to the *even dumber* "techbros insisting everyone trust the Machine That Gets Stuff Wrong to do the thinking for them" genius. how long are we thinking til the first big data breach caused by shitty AI "vibe coding"?