Argh so many great ideas to add to the list!

But the great news is that we've just about completed brining on a long-needed developer, so more features should be appearing soon!

While pot-smoking Elon has been wasting his time recycling plastic bottle rockets, we've succeeded in a major breakthrough: clicking "markdown" will now open a separate tab, so there's no danger of any text loss for anyone, and no more loss of spectacular arguments.

@TamayBesiroglu wrote:

See jkaufman's google doc on probability estimates of cryonics success for several good factorisations from various people:…

Very nice. Perhaps we should make a cryonics series at some point.

@Fruo @bzial. Oops! Totally glossed over the second part of the resolution. Apologies, I guess I'm still jetlagged from a trip! Unresolved.

There's a super in-depth 4-part report from Ajeya Cotra at (/sponsored by?) OPP on transformative AI timelines. [Part I]( gives an overview. In super-brief, she uses four different hypothesis on which to ground estimates (in the form of pdfs) for the total FLOPs of computation required to train something roughly as effective as a human. p. 14 shows the assembled pdf for this. It's very broad: > the distribution places non-trivial probability mass on a range of 26 o...

@frxtz Spam report button and @moderators are both coming very soon!

@(Uncle Jeff) wrote: > Am I reading this right, that no negative resolution is possible before 2100? Since living long enough to see a negative result would itself produce a positive result, I know which way my prediction is going. > > And then there's the implied meta-question: Can Metaculus itself survive to resolve this question negatively? This is a common question. We've gotten definitive proof by tachyon transmission that Metaculus will be around well into the 22nd century, and has (by that time) grown vast in importance such that the people mak...
So I've been scanning over reaction to the Alphastar win(s) and blog. Implementing the actual resolution listed in the question is hard and I'd like to suggest that this resolves positive. There is definitely a good amount of discussion as to exactly how impressive the win was, given that the AI could make extremely precise movements at a peak rate higher than humanly possible, and that it had totally situational awareness rather than being confined to a particular view that it had to change. See for example [this hacker news thread](https://news.ycom...

So far, the community score is crushing it, with an average log score of 0.035 over 10 binary questions. I don't expect that to last, but it's encouraging.

Also much better than the Metaculus prediction, which we need to look into.

In my view, if Biden wins, then Metaculus has a good score on this question, and if not he doesn't. The fact that it's a close race does not come in, IMO. The reason the Trump probability was low here and elsewhere is that *even with* a significant polling error, Biden was still on-track to win. A significant polling error happened, and he is (as of this writing) still on track to win. If you want to score the fact that it's a close race rather than a more decisive one, the more appropriate place is at [this question](

I'm updating to a significant probability here because (a) he's now leaving office; (b) it seems very plausible that he's broken a lot of laws that can be prosecuted at the state and not just federal level; (c) precedent: Metaculus has historically been surprised by people in the Trump administration going to prison.

After looking at all the below and the comments online, I feel pretty persuaded like @AngraMainyu and @traviswfisher and @gjm and others that (a) the comment would count toward resolution, and (b) it's 95+% likely to be Gowers that made it. (The idea that some highly-knowledgable mathematician is impersonating Gowers under Steckles and others (and Gowers') nose seems pretty farfetched.) Since we're not giving out a Fields medal or a Clay Prize here but just some points, it seems to me like this is enough so unless I hear violent objections I'll unresol...

It has probably occurred to some predictors that there is an opportunity to gain lots of points by predicting a large number for this question, then making a contribution so as to cause that prediction to come to pass. I just want to assure everyone that this is perfectly acceptable.

PSA for forecasters to pay careful attention to the tails of your distributions. The data being solicited from the experts is in may cases 5%, 50% and 95% bands (NOT the 25-50-75 displayed on Metaculus), so be sure to use the CDF readout to make sure that you like the values your PDF is giving for these since they will be directly compared.

Imagine, for example, chatting with this thing. Judge: I live in a two story home with the kitchen on the second floor. I don't have a coffee machine, but one of those press-pot things, a metal one. I can't recall if the coffee is in the fridge or freezer. There's a plot in water-heater-thing. Now please describe for me, in step by step detail, the procedure by which you would make a cup of coffee (with cream and sugar) after walking into my front door. Human: I want in the door, and find the stairs to go upstairs. I find the kitchen and go in. I lo...

Looks like the comments above are that we will do what Amherst does. Strongly agree we should have this clarified by all parties before the next round.

I see a strong case for resolving at 959,056 (since this was the intention of Amherst, this was noted in another discussion, and it's in the spirit of the question.)

I see a pretty strong case, based on Metaculus precedent of being very literal with resolutions, for resolving at 931,698.

That makes a quite strong case of ambiguous resolution, or maybe even interpolating (i.e. splitting the difference.)

I'm going out on a limb here with a high prediction. Everyone says this is going to take 12-18 months, but (a) I don't think we've seen anything like this degree of urgency in the history of modern vaccines; (b) There really are very fast advances in biotech; (c) There are highly-affected countries, in particular but not just China, who will be willing to take a higher risk with early test subjects be able to push through a faster review process. It would be interesting to understand from first principles (rather than precedent) what the minimal ti...

@emilowk Your crass scheme of throwing money at the problem is totally working, as this has been bumped up the queue and will probably get taken up in the next week or so.