History of Science & Technology Q&A:
History of Science and Technology Q&A (April 12, 2023) »
1 hour 6 minutes
✕ close
Do you think it will be possible to recreate historical figures as bots to interact with and get their perspective on current research areas?
Why do many great mathematicians complete their most influential work in their early 20s?
Does "prompting" (as for LLMs) have some historical precursors?
So Feynman could have been a great prompt engineer (given that he was such a great expositor/teacher)?
How do you think future researchers will look back at this current time in history? We look at bones and architecture to determine facts about the past; what will they look at to determine facts of our time? AI?
Can we restore old, lost books by reading other old books which talk about them?
Seneca wrote many many letters. Could we detect if some have been wrongly attributed to him?
I love a historian David Lewis's possible world that we can create alternative history/hypothetical situations to learn what went wrong historically. I just wonder whether AI can utilize deep learning to generate the sequence of historical events with the constraint of data and and recreate the alternative historical events with the known variables to generate hypothetical outcome?
Isn't sonographic/x-raying safer than digging through ancient architecture? Or is it still dangerous somehow?
They have been using muons to probe the pyramids in Egypt.
Maybe AI can help with such more passive imaging through buildings?
Neural network weights will be a more efficient means of archive through the centuries than books and libraries—which will matter as with ChatGPT the volume of published writing will climb exponentially.
Prompting has relevance in psychology and philosophy.
Could it be that the best prompter now are poets? Or better... computational poets? I don't think RAM or ROM-chips will survive the passage of time or solid state drives... View Less »