alt Hacker Newsit's not going to happen with LLMs unless ram + storage gets several orders of magnitude cheaper like, yesterday
informatics aren't magic, you'll never be able to compress """knowledge""" into a small model in a way equivalent to the 1.5 TB model
Replies
acidhousemcnab • yesterday at 10:38 AM
This will happen, but reconfiguring the infrastructure of the entire planet to train LLMs and run them over networks might be the "bubble", the megalomania.
I agree. But I also think the future is some kind of hybrid approach where agents run locally, what they can, and then call out to the cloud for what they can't.