@CoinDesk@IOHK_Charles really screwed up with that statement, and he'll come back and say that's not what he said and the media is bad, obviously he wanted a different effect, but that's how it turned out
😅
@PINTO03091 may i ask you, how to deal with manufacturing data, a lot of lines, from different sensors, one central file, is it possible to make some prediction for example what is the best way based on historical data. i know the is shadow mode, is it possible train the model?
@theo i am building enterprise RAG, i switched from gpt-5.4-mini to > gemini-2.5 and result went from 60% to 80%, just switching model. I do not know but i seems to signal landed better between dots in black box for technical manuals