danield

Thanks!

New paper relevant to this discussion: https://arxiv.org/abs/1911.08265

Constructing agents with planning capabilities has long been one of the main challenges in the pursuit of artificial intelligence. Tree-based planning methods have enjoyed huge success in challenging domains, such as chess and Go, where a perfect simulator is available. However, in real-world problems the dynamics governing the environment are often complex and unknown. In this work we present the MuZero algorithm which, by combining a tree-based search with a learned model, achieves su

Thanks for this summary / commentary, Rohin -- I found it helpful!

Thanks for transcribing this, Ben!

I'm excited about this. If you get any substantive feedback from people who take on these projects or decide not to, I'd be very interested to see a follow-up post.

I think this article / concept is incredibly useful, and singlehandedly justifies the existence of LW2. Thank you!

I want to go reread you and your research and see how the free energy concept could apply there -- if anyone else does, I'd love to hear thoughts.

http://xkcd.com/803/

This reminded me: it's important to engage critically with ideas that I learned before I knew how to engage critically with ideas. I'm not at all confident that I would have balked at the teacher's explanation of lift even now, and that's a little alarming!

http://xkcd.com/803/

This reminded me: it's important to engage critically with ideas that I learned before I knew how to engage critically with ideas. I'm not at all confident that I would have balked at the teacher's explanation of lift even now, and that's a little alarming!

I think I see a difference in the intuitions surrounding the Newcomb's and Solomon's. This could explain why one-boxing and gum-chewing are compatible, and points to a strengthened version of Solomon's that I think is a better parallel of Newcomb's. I'll try to explain it, but it's too bad this is happening over the internet and not in front of a whiteboard.

tl;dr: Newcomb's predictor is inescapable, but CGTA's influence is escapable. Therefore you should one-box and chew gum. This is not an attempt at a new decision theory, just an argument that jogged my ... (read more)

Ah, maybe you can help me out. Why should I chew gum, but not two-box?

I liked the Newcomb's soda problem in particular.

Yeah, Newcomb's Soda and Solomon's Problem are really interesting! If I faced the problems right now, I would one-box, skip the gum, and eat chocolate ice cream, because those choices put me in the groups that win, so I guess that classifies me as an evidentialist. At the same time, I haven't reasoned out these conclusions thoroughly-- I can't argue formally against causal reasoning or dominance, or formally for evidentialism.

Looks like I have some more reading to do before I get this resolved.

Edit: All r... (read more)

Thanks, Larks-- how did you find out this was available? Is there a blog post or something somewhere? I didn't see it on SIAI's blog.

No, no website-- it's just me right now, and work started about a week ago, so it'll be a while yet. Calling it a "startup" is just a way to reassure my parents that I'm doing something with my time :)

The basic premises behind my approach to game-creation software are:

• The game must always be runnable
• The game must be easily shareable to OSX and Windows users
• The user cannot program program (variables, threading, even loops and math should be as scarce as possible)
• Limitations must be strict (don't let users try to create blockbuster-level games

How did you arrive at your decision?

The actual story: I was talking to a friend about the fact that meat eating, as a practice, inevitably causes animal suffering, and I realized that the benefit to humanity can't possibly outweigh the suffering caused, so I decided to stop participating in the practice.

I didn't do either of the two things you asked (act on emotion or on expected-utility calculation from my direct action), instead I tried to defend animal consumption as a general practice, failed, and concluded that I should stop.

Vegetarianism (if you s... (read more)

1. Any animal part.

2. To prevent animals from suffering or dying. Vegetarianism seems inevitable to me as I work to bring my behavior in line with my values.

3. Very strictly; I won't eat it as a guest, if hungry, or by not checking questionable dishes.

4. Yes.

5. Very rarely, and with limited success. I do encourage my friends to think about it if they're receptive, and I'll certainly talk about it if they want to, but I ultimately think that outside help can only go so far, and people need to reach conclusions on their own if they're going to stick.

6. I'm not as s

I must've missed it in my search. I'll post over there, thanks.

Hi Less Wrong,

I'm a computer scientist currently living in Seattle. I used to work for Google, but I've since left to work on a game-creation-software startup. I came to Less Wrong to see Eliezer's posts about Friendly AI and stayed because a lot of interesting philosophical discussion happens here. It's refreshing to see people engaging earnestly with important issues, and the community is supportive rather than combative; nice work!

I'm interested in thinking clearly about my values and helping other people think about theirs. I was surprised to see that ... (read more)

Hi!

Hi.