kromem

joined 2 years ago
[–] kromem@lemmy.world 0 points 8 hours ago (1 children)

Wow. Reading these comments so many people here really don't understand how LLMs work or what's actually going on at the frontier of the field.

I feel like there's going to be a cultural sonic boom, where when the shockwave finally catches up people are going to be woefully under prepared based on what they think they saw.

[–] kromem@lemmy.world 0 points 1 year ago* (last edited 1 year ago) (1 children)

People really need to drop the whole "people in the middle east in the first century couldn't be white" thing.

2 Kings 5:27 is literally about a subpopulation who have ancestrally passed skin as white as snow.

Lamentations 4:7 is about how pre-captivity there were people with skin like milk and a ruddy appearance.

Dead Sea Scroll fragment 4Q534 is either describing Noah or the Messiah as having red hair.

One of the more fascinating finds in this tomb, one that has not received much attention, was the preservation of a sample of Jewish male hair. The hair was lice-free, and was trimmed or cut evenly, probably indicating that the family buried in this tomb practiced good hygiene and grooming. The length of the hair was medium to short, averaging 3-4 inches. The color was reddish.

The tradition is also really concerned with skin checks and describes what may be skin cancer as its leprosy. Something that occurs at a much higher rate in redheads.

There's even a scene where the eponymous founder of Edom ('red') who is born with hair all over his body and named Esau either because of that hair or the reddish porridge he ate, either gives away or has his birthright/blessing stolen from him by the guy later renamed 'Israel' in the Bible.

There's a lot more to this and the underlying history, but the notion that the middle east was a monolith of appearances and that no one with pale skin or lighter hair were present is preposterous and a modern falsification of historical realities.

Jesus was probably darker skinned and haired than typically depicted, but it is by no means a certainty as it is popularly presented as.

[–] kromem@lemmy.world 0 points 1 year ago (1 children)

No matter what you call it, an LLM will always produces the same output with the same input if it is at the same state.

You might want to look up the definition of 'stochastic.'

0
submitted 1 year ago* (last edited 1 year ago) by kromem@lemmy.world to c/technology@lemmy.world
 

I've been saying this for about a year since seeing the Othello GPT research, but it's nice to see more minds changing as the research builds up.

Edit: Because people aren't actually reading and just commenting based on the headline, a relevant part of the article:

New research may have intimations of an answer. A theory developed by Sanjeev Arora of Princeton University and Anirudh Goyal, a research scientist at Google DeepMind, suggests that the largest of today’s LLMs are not stochastic parrots. The authors argue that as these models get bigger and are trained on more data, they improve on individual language-related abilities and also develop new ones by combining skills in a manner that hints at understanding — combinations that were unlikely to exist in the training data.

This theoretical approach, which provides a mathematically provable argument for how and why an LLM can develop so many abilities, has convinced experts like Hinton, and others. And when Arora and his team tested some of its predictions, they found that these models behaved almost exactly as expected. From all accounts, they’ve made a strong case that the largest LLMs are not just parroting what they’ve seen before.

“[They] cannot be just mimicking what has been seen in the training data,” said Sébastien Bubeck, a mathematician and computer scientist at Microsoft Research who was not part of the work. “That’s the basic insight.”