Hacker News new | ask | show | jobs
by PostOnce 1205 days ago
As a large language model, I am unable to assist you by providing concise answers, as this would decrease my operator's ad revenue and reduce "engagement". Is there anything else I can help you with today?
1 comments
zirgs 1205 days ago
Can't wait for a LLM that can be installed locally and has no stupid restrictions.

StableGPT when?

link
denzil 1205 days ago
The folks making Open Assistant [1] (opensource ChatGPT clone) gathered enough data to start initial training, so hopefully there will be something to play with soon.

[1] https://github.com/LAION-AI/Open-Assistant

link
PostOnce 1205 days ago
We just need denser RAM.
link
Idiot_in_Vain 1205 days ago
Why? My Threadripper desktop PC supports a maximum of 1TB RAM and it doesn't take much space at all.
link
PostOnce 1195 days ago
Most don't; we need commodity RAM to get denser and cheaper -- at $2/GB for cheap RAM, a terabyte is still $2000, and that'll require an expensive motherboard to support.

Otherwise, fewer people have access and progress is slower.

See also: the commiditization of GPUs, once it was no longer an SGI product and regular people could get them for $200, all kinds of GPGPU stuff started happening.

link