So the meetup place will be at my apartment again, on Saturday 8th from 1-4 pm:

66/B/1 Ulloi Ut
6th floor, door 3
Ring 27

Sorry for not having a meeting in January, I've been a bit busy with income generating projects + taking care of the toddler.

We're going to be talking next time about the question of if it makes sense for us to put point estimates on the probabilities and then use them for expected value calculations for events that are not drawing from a clearly understood probability distribution.

Does it make sense to say your P(Doom) is 50%, or should you be saying that it is between 10 and 90 percent? Or should you be saying that you have no idea, but you are fairly sure that it is plausible that things could go well and that we might all die without using any numbers at all?

Here is an essay talking about the general subject. It is rather long and technical. This subject has long interested Milan, and hopefully we can collaboratively explore this idea space. To highlight one thing from the article: Milan is quite interested in if the article's concept of "maximality" makes sense and if it's practical to use. If some of you might want to argue for the (seemingly much more common in LW circles) classical bayesian position of preferring point estimates (or any other positions), we can potentially have a nice and lively conversation.

Article in focus: https://www.lesswrong.com/posts/zMCHujoAdgiCYo7JN/should-you-go-with-your-best-guess-against-precise

Some further reading material for those who are especially interested in the general subject: 
https://plato.stanford.edu/entries/imprecise-probabilities/
https://sipta.org/info/documentation/
 

Hope to see lots of you there,

Tim Underwood