clueless

Given the current llm context size limitation, what is the state of art for feeding large doc/text blobs into llm for accurate processing?

this article has thrown me into a rabbit holeee.. what a fascinating history. thank you astralcodexten

This all seems to be the precursor to the recently viral Matthew McConaughey video saying he wants a private LLM, and to think, first of all, we still don't have a great/mainstream tool/product for this, and second of all, it's more than 80 years in the making, is mind boggling...

hmm I'm confused, whats the sell with inaa compared to vlc?

more than 15 years ago I got a chance to play with gnu radio and back then it was hailed as the next big industry.. fast forward, and beyond the hacking community (and the hobbyist), it still has not taken over.

ah it's not? that would be a blocker for me

and then... "Many of the people involved in the mission were later promoted."

that's not been my experience with the "discover weekly" feature

>They found no guns or uniforms. Evidence suggested that the crew, which people briefed on the mission said numbered two or three people, had been civilians diving for shellfish. All were dead, including the man in the water. Officials familiar with the mission said the SEALs pulled the bodies into the water to hide them from the North Korean authorities. One added that the SEALs punctured the boat crew’s lungs with knives to make sure their bodies would sink.

Nothing to see here but a bunch of psychopaths killing innocent people as they screw up their own mission

really the biggest service spotify has for me is its music recommendation engine... and so the big question with all this setup is: is listenbrainz's recommendation engine better than spotify?

what are some hardware people would recommend (specifically looking for power efficiency)?

dang, looks all sold out online. Anyone have any leads? edit: looks like about $225 shipped from german amazon store

woah, yeah, what is up with the galaxies spin motion/transformation? there's gotta be a good explanation

I'm curious, what was the podcast episode?

this is not the same technology as e-ink, hence why they call it e-paper

Elon did not coin this, Kurzweil has been using this coinage for a lot longer.

ah in that case, this is long over due.

does such an open source project exist?

in an open access world, who should be paying the people who are reviewing the papers (which cost time/money)?

that's a great concern as I'm in the same boat. What have you find with regards to the monitoring/security solution?

So what are the best alternatives?

how are "trillions of humans living in the far distant future" improbable, or poor information or has a lack of information? seems pretty obvious

the result of the last World cup (2022) finale being the absolute best example of such engaging match...

"Hibiki currently only supports French-to-English translation."

Sure, openly written documents can help with evaluation of intent, but how can we ever define someone's intent (something that is only in people's hearts)? We know of many legal cases where the intent is obvious but not easily provable.

so let's use up those extra CPU cycles and update the UI to slow everything down again.

Precisely, I'm not sure why Curtis is kept being discussed when if you listened to him now, it's all nonsense.

spoiler alert: is copycat strategy the reason united states feels like it's becoming more like the rest of the world, more authoritarian?

This stuff always reminds me of "The Adventure of the Greek Interpreter" from The Adventures of Sherlock Holmes TV series (This is Season 2, Episode 2, originally aired in 1985). In this episode Sherlock Holmes and Dr. Watson are at 221B Baker Street, and Holmes gives a detailed demonstration of his observational and deductive skills by people-watching out the window.

Ok, following the following comment in this thread fixed the issue: https://news.ycombinator.com/item?id=43943624