
147. Large language models (LLMs) - interview with Dr Guy Emerson
Education Bookcast
The Problem With Scientific Training Data
The testing doesn't need to match necessarily the training so the my understanding of the the point of that paved the wall you know semi-paper whatever you want to put it right. If they don't tell us the training data how do we know that's true right? It becomes impossible to verify that claim and i think that's extremely problematic if you have a scientific paper where it's impossible to verify the claims of the paper that's why i don't see it is scientific, he says.
00:00
Transcript
Play full episode
Remember Everything You Learn from Podcasts
Save insights instantly, chat with episodes, and build lasting knowledge - all powered by AI.