This is from a conversation I had with a friend:
- Twitter: Generally, I think Twitter is a good place for news AND expert knowledge, since there are many self-doxxed experts using twitter as a personal branding-cum-networking media in a much more casual and honest setting versus LinkedIn. You just have to curate the right list of follows to make the recommendation AI work for you, rather than just for itself
- Things to try:
- Having multiple twitter accounts following different groups, such as for macro, for crypto, for machine learning and data science
- Things to try:
- Reddit: Decent too with some great communities, better for long-form knowledge sharing than Twitter. However I think with Reddit it’s harder to curate since it’s based on interest groups rather than follows.
- Metaculus: A recent discovered gem. I love it because people on the platform are selectively versed in practising Bayesian Updates mental model. I follow certain “bets” that are of interest to me, for example I was following https://www.metaculus.com/questions/8757/omicron-variant-deadlier-than-delta/. The comment section is REALLY good since people share pretty reliable news to explain their bayesian updates for adjustments to bet probabilities, so it serves as an excellent aggregator of reliable news for me.
- LessWrong forum: OG Rationalism Bayesianism
Daven Lim 