wills_forward

stick enough floppy's in parallel and you could do the same thing

neat

There's a lot of opportunity in being the manager who can still see it

Why not use both? I just built a pipeline for document data extraction that uses PaddleOCR, then Gemini 3 to check + fix errors. It gets close to 99.9% on extraction from financial statements finally on par with humans.

The cheap easy take: it's tragically ironic that the software running the infrastructure in Silicon Valley is such a problem

So this could universally decrease the memory requirements by un-quantitized LLMs by 30%? Seems big if true.

This paper is basically statistical mechanics with a quantum veneer. Two major issues:

1. Scale: They're simulating just 13 qubits with QuTiP and making grand claims about quantum thermodynamics. The computational complexity they're glossing over here is astronomical. Anyone who's actually worked with quantum systems knows you can't just handwave away the scaling problems.

2. Measurement Problem: Their whole argument about instantaneous vs time-averaged measurements is just repackaging the quantum measurement problem without actually solving anything. They're doing the same philosophical shell game that every "breakthrough" quantum paper does by moving around where they put the observer and pretending they've discovered something profound.

It was funny to hear the same guy warning LMMs were getting too powerful now talking about the limits of available original training data.

Citadel or Jump?

I really like the elegant simplicity of tagging the screen elements like that and not obfuscating it away.

Nice work too!

Is this part of the reason Apple decided to support RCS? They knew the iMessage system would get opened up eventually anyway...

Aw thanks for such encouragement all

My jaw drop to see algorhythmic complexity laid out so clearly in a 3d space like that. I wish I was smart enough to know if it's accurate or not.

This seems like a big step forward in terms of running specific use-case trained inference locally, right? At least given current hardware generally deployed in business.

Great podcast

MIT and Microsoft Researchers Introduce "RetNet" - An 8X Faster Transformer Alternative for AI

This is such a bummer, but not shocked they're out-of-bounds given the bullets in the letter and how big they've become: https://www.cftc.gov/csl/22-08/download

Does anyone see explainability as another good reason to trees on tabular data, for which I think users would expect more digestable outputs?

That Twitter thread made me MORE of a believer that Dall-E has a language its own. As others said, seems like the argument is more about defining "language".

Because it will legitimize crypto, not discourage it.

I’ve been looking for exactly this. Signed up for premium after doing my first model. So well built, very nice work.

“When he and his entourage flew back to Georgia on their private jet, we were like, what happened today,” Malone said. “He probably was sitting in that plane and was like, ‘what a weird place.'”