3 ms·
if everyone is using LLMs to solve problems, in a few years, won't LLMs run out of content to mine? In short, how can the general dumbing down of LLMs and degra
by notepad0x90 2y ago
if everyone is using LLMs to solve problems, in a few years, won't LLMs run out of content to mine? In short, how can the general dumbing down of LLMs and degradation of content used to solve problems be avoided over the long term?
For questions about events and problems that arose after 2025, where would LLMs get information to solve those? and who would be asking those at a forum LLMs can access going forward?
Is the snake eating it's own tail?
- kadushka 2y ago1. People will continue to answer questions and post about events and problems after 2025. Eventually LLMs themselves (inside robots) will be observing the world and reporting on anything interesting. 2. Best LLMs today answer questions better than 90% of people who comment on forums. So if these LLMs have been able to train on all the crap posted on internet so far, they should only get better as they are being trained on high quality output from the latest (and future) LLMs.
- visarga 2y agoMost interesting data will be collected in chat rooms and apps. There are over a billion LLM users, they act as human-in-the-loop enhancig the LLM, sometimes testing ideas even in reality. Wondering what providers are doing with the chat logs.