Live data from Hacker News

The Coming of Local LLMs

nickarner.com

211–214 of 214 posts

Re: The Coming of Local LLMs

#211
post #196

Earlier quoted context omitted.

The first step is for it to be viable for smaller models on desktops. The rest will follow, as hardware catches up. Hybrid analog-hybrid NN hardware is on the horizon, maybe in 3-4 years. This would allow GPT-4 level performance on an iPhone with plausible battery life. The current hardware of course can't pull anything like this yet. But iPhone supports on-device facial recognition, object recognition, dictation and…

Who is making said hardware?

There are few efforts under way.

https://research.ibm.com/blog/why-we-need-analog-AI-hardware

https://news.mit.edu/2022/analog-deep-learning-ai-computing-...

Re: The Coming of Local LLMs

#212
post #45

Earlier quoted context omitted.

LLMs as general purpose classifiers is a really big deal, especially because you can give them fuzzy instructions. I know people are worried about LLMs and spam, but I think LLMs may provide an opportunity to elevate online discourse by being more efficient at filtering out spam and low quality commentary.

I already have a custom browser plugin that calls out to GPT (gpt-3.5-turbo is cheap and good enough for this) to classify and filter out low-effort, overly negative, or intellectually dishonest HN comments. It significantly improves the experience on this site. Bonus points: I had never written a browser plugin, but GPT4 helped me do it in under half an hour.

[flagged]

Re: The Coming of Local LLMs

#214

Were transformers used in other contexts ? biochemistry ? geometry .. whatever.

Alphafold uses something they call Evoformer, it is an attention mechanism. Our group has tried, and so far failed to utilize transformers for a very very specific search problem in geometry ( https://bit.ly/unit-distances ).

thanks a lot man, very nice
Post reply on HN