AI Forecasting and AI for Forecasting

I'll give a high level overview of my thoughts on AI forecasting (including our next big scenario), and on how we should use AIs for forecasting (mostly AIs more capable than today's).

The latter will be more broadly about AI for epistemics rather than about forecasting narrowly, but "AI Forecasting and AI for Epistemics" isn't as good of a talk title.

Then I'll do a Q&A.

(subject of talk might change)