adamnemecek

> For years, AI researchers have used math as a proving ground for their models.

For years?

I feel like it's harder to hide malicious stuff in Rust build scripts.

Autodiff is preventing any meaningful discussion about safety, systems trained with autodiff cannot be made safe.

Here's the repo link https://github.com/NVlabs/cuda-oxide

This is exactly how the open sourcing of Swift went so I imagine it will be the same.

+1 for Typst being amazing.

I can actually like write my own functions when I need to. I don't think I have ever written a LaTeX macro without having to look up a lot of stuff.

That's like saying that Ted Kaczynski was innocent, because he didn't force anyone to open the packages.

They are back because modern languages (Rust, Go) have made it pretty straight forward to build them. Ratatui and such allow you to write a TUI really quickly without needing to deal with VT100 arcana.

I think that you send the entire conversation with every request.

It really doesn't, Rust is a better language.

It makes complete sense to polish that usecase.

This blog post is less than excellent. In fact it's a bit of a shit.

Neither did Steve Jobs and no one cared.

What do people use for Windows-like window management on macos? I tried a bunch of them and I'm not a fan of any of them.

I actively dislike the notion of spaces.

Wow, fresh take. Not.

Good call, ArXiv seems like one of the most important institutions out there right now.

Who could have predicted that a guy who has no experience developing superintelligence will fail at developing superintelligence.

Gradually develop an OS that is not just an Android fork, but a full blown OS people can contribute to. And of course write it in Rust, like the problems with Java are so apparent in Android.

Robots or drones with ground penetrating radar?

This feels like a perfect use case for AI.

It might be a good idea to look into the research on streams as coalgebras, there is quite a bit, for example here https://cs.ru.nl/~jrot/CTC20/.

Coalgebras might seem too academic but so were monads at some point and now they are everywhere.

Did consider that your view might be skewed because you work in a CRUD app?

That is very sad to hear.

But you have to know what book you are looking for.

The previews are still there though, they just don't rank.

This is actually pretty good.

But why would people train on excerpts from Google Books when whole books can be downloaded on libgen and such?

No the search results went from pretty good to absolute garbage https://bsky.app/profile/adamnemecek.bsky.social/post/3mdbup...

The books are still there, it seems like the rankings have changed though.

The corpus is still the same, like searching the name of the book will find it, but the full text search.