Hacker Newsnew | past | comments | ask | show | jobs | submit | someothherguyy's commentslogin

then it would cost human lives, no more fun i made this thing with no real effort vibes

> but at least it sounds theoretically doable today

why


We have a bad universal classifier now (via Jev). 0->1, one might say.

A bad universal classifier does suggest a good one later. And that is exactly what I would call "theoretically doable"

That said, I don't think that Jev is a magic breakthrough or anything. I think it is just a particularly good narrative with an easy way to try it out.


LLMs are like lossy compression of ~all of written text ever produced, with useful recall. To the extent that the corpus contains labelled examples of the given classification task, it's not unreasonable to think that we'll be able to build a decoder for that, just like we already have a useful decoder for next-token prediction. Extend to image classification the same way we already have multimodal LLMs.

any instances of this in the wild? it seems like fridges are pretty cheap anyway and ones with smart features are more expensive? i don't think this a tv-like scenario.


https://archive.ph/U3CVN

Cant manage to get an archive link for the second one.

Does anyone know of an xcancel like service but for Reddit? OP links just lead to a login page.


Redlib works. Either self host or use a public instance: https://github.com/redlib-org/redlib#instances

They now require login for old.reddit.com. You can remove the subdomain and access the modern JS laden slop without an account unless they deem it 'Adult Content'

Reddit now requires login. What are these?

There was a fake story that went viral last year about a woman being sent to the psych ward after seeing the Pluribus ads on her Samsung fridge.

The ads were real though: https://ibb.co/5hbG7ymS


> Occam's razor: they couldn't make any more substantial improvements.

doesn't sound like a razor at all



https://scholar.google.com/scholar?oi=gsb20&q=Mars%20Sample%...

That paper outlines a plan to do it, which people have been trying to do since the 70s

Here is one that will launch soon: https://en.wikipedia.org/wiki/Tianwen-3


> The mission consists of 15 different vehicles and spacecraft plus two launch vehicles

Yeah still looks like an overcomplicated project

Maybe that's why China can make it work


too powerful to give up, sweet imagick love

you can't in imagine the same way you couldn't imagine generating the text from your initial prompt. when will the hubris end?

some variant of "oh i am capable of generating text i know nothing about, but it helps to have a human i can rely on to tell me more about the text i know nothing about."

pick up a bachelors degree worth of books to humble yourself


https://www.nature.com/articles/s41586-026-11032-2

Developmental xenocortication using human-derived organoids in mice


What? You don't like horizontally scrolling 6000 columns to read something?

https://en.wikipedia.org/wiki/Code_folding


Yeah, if IDE's allowed code folding of comments then the authors original problem (having to scroll past comments they thought weren't useful) wouldn't be that big a deal.

many do. for example: https://www.jetbrains.com/help/idea/code-folding-settings.ht...

there are also extensions for vscode that do this, vim plugins, etc


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: