Wiki Contributions


Was extremely enjoyable.  Any plans for more?    

Huh.  I found your forecasting newsletter via LessWrong, and then subscribed to the substack's RSS feed?  Which probably made me less likely to open it/see it in LessWrong?  Dunno.  Maybe your LessWrong traffic moved to substack?  (sample size=1)    

Yeah.  Why I said "generally."  Obviously if you emulate Stalin bad things will happen.  Slightly off topic but I'd say those leader's weren't "unconstrained" and revealed something about human nature, but were constrained as any ruler and had to follow the Dictator's Handbook to stay in power.  

Move to New Hampshire.  Join the Free State Project.  I'm only slightly kidding.  For national politics, a huge amount of people dedicate ginormous amounts of attention too it and because of that there's not much slack for an individual person to affect much.  Local politics on the other hand, people barely care about.  Pull the rope sideways and all that. (   Apparently, something like 3000 libertarians have been able to accomplish all this:  just this year.  The movement has elected 40 of their own members to the senate.  (Out of 300ish)  So a sizable minority that is vocal can have a huge impact on local politics. And furthermore, another 1000 have moved to NH year due to Covid.  So if you have libertarian leanings (and/or think those policies are good at helping people) and want to run for office , I'm sure they would get you elected pretty quickly.  Apparently some seats run completely unopposed, just because how few people care.  So yeah, if you want to have a meaningful effect, on a small(ish) amount of people, then move to NH.  

Silly question warning. 


You think that when an AI performs a bad action, (say remove the diamond) the AI has to have knowledge that the diamond is in fact no longer there.  Even when the camera shows the diamond is (falsely) there and the human confirms that the diamond is there. 

You call this ELK

You want the human to have access to this knowledge, as this is useful to choosing decisions that the human wants.

This is hard.  So you have people propose how to do this.  

And then people try to explain why that strategy wouldn't work.

Rinse and repeat.

Once you have a proposal that nobody is able to show doesn't work.... profit?


Correct any misunderstandings in my basic overview above.  

Continuing with the pattern, our next meetup would be Saturday, Nov 13. @Connor /CitizenTen won't be able to make it, so Zach offered to help coordinate. We'll meet at the same place as last time: Bird Park (here, at least initially: and the same time: 1pm PT. 

Separately, Zach volunteered to help put together a small activity to try to encourage interesting conversations. We can pilot it at this next meetup, if there's interest. (Like any pilot, we can bail on it if it doesn't seem to be working.) Elevator pitch: A game where you answer questions to show how well you know someone — except that you just met them, and have only had 10 minutes to talk. The idea here is to encourage folks to cut through some of the "small talk" and try to figure out how to quickly/efficiently learn enough about someone that you can predict how they might answer some questions. The rough idea is that we'll partner up, get to know each other, and then have some of the partnerships answer questions. • To assess how well you understand your partner's values, I might ask: Does Partner [A|B] believe torture is ever permissible? • To assess how well you understand the way your partner evaluates trade-offs, I might ask: Partner [A|B] is offered a choice between $10k and a 1-in-10 chance of $100k. Which do they pick? • As a fun experiment in coordination, I might ask: Pick a number. Of course, I wouldn't ask exactly these questions; that would be too easy. If you'd like to suggest questions, we can build a pool: (don't share them here to avoid spoilers).

Hmmm... It seems that someone has thought of doing the exact same thing? Though considering the digging I had to go to find it, it should be more prominent. 

On the having a shoulder Duncan, I explicitly tried to do this upon listening to you being interviewed by Spencer Greenberg.  Sadly, digital Duncan doesn't exist in large enough quantities to emulated.  So for those of us that don't happen to live (I assume) in the Bay Area, can Duncan increase how many podcast or video's he's on?  Having a shoulder Duncan sounds really useful to a large number of people you interact with and enlarging that pool purposely seems to be a pretty good idea. Food for thought?

Thanks for the help! Now let's just hope that no one reads it as 8 pm and decides that's the time!

Load More