Machine learning and large language models will be very important ingredients in future science, but the hype around them (claiming they will solve climate, poverty, and disease) represents 'simplemindedness' and 'Dr. Pangloss speaking English'—applying reductionist thinking to complex multifactorial problems.
normativepending
Speaker
David KrakowEvidence Quote
“not to say that machine learning and large language model won't be a very important ingredient in all future science that's true right but this is simplemindedness this is Dr pangloss speaking English and actually there's a very nice paper by uh Terry saginowski at the sulk on what's called uh the reverse churing test that the intelligence of the machine is a mirror of the intelligence of the interlocutor right that if in other words um if you ask clever questions it gives clever answers if you ask stupid questions it gives stupid answers so what llms really are is mirrors”
Created: 8/10/2026, 10:52:22 PM
My Notes
Loading notes...