User Profile

star3059
description127
message5028

Recent Posts

Curated Posts
starCurated - Recent, high quality posts selected by the LessWrong moderation team.
rss_feed Create an RSS Feed
Frontpage Posts
Posts meeting our frontpage guidelines: • interesting, insightful, useful • aim to explain, not to persuade • avoid meta discussion • relevant to people whether or not they are involved with the LessWrong community.
(includes curated content and frontpage posts)
rss_feed Create an RSS Feed
All Posts
personIncludes personal and meta blogposts (as well as curated and frontpage).
rss_feed Create an RSS Feed

Announcement: AI alignment prize round 2 winners and next round

1mo
1 min read
Show Highlightsubdirectory_arrow_left
22

Using the universal prior for logical uncertainty (retracted)

3mo
1 min read
Show Highlightsubdirectory_arrow_left
13

UDT as a Nash Equilibrium

4mo
1 min read
Show Highlightsubdirectory_arrow_left
17

Beware arguments from possibility

4mo
1 min read
Show Highlightsubdirectory_arrow_left
14

An experiment

4mo
1 min read
Show Highlightsubdirectory_arrow_left
10

Biological humans and the rising tide of AI

4mo
1 min read
Show Highlightsubdirectory_arrow_left
20

A simpler way to think about positive test bias

4mo
1 min read
Show Highlightsubdirectory_arrow_left
9

Beware of black boxes in AI alignment research

4mo
1 min read
Show Highlightsubdirectory_arrow_left
10

Announcement: AI alignment prize winners and next round

4mo
1 min read
Show Highlightsubdirectory_arrow_left
68

Announcing the AI Alignment Prize

7mo
1 min read
Show Highlightsubdirectory_arrow_left
79

Recent Comments

Another analogy is that Bayesianism = decision theory and frequentism = game theory. Bayesians play the best response to a particular mixed strategy of nature (prior), while frequentists try to achieve "frequentist guarantees" which is like minimaxing against nature.

One possible problem with the scheme is that level N assistants will have access to a lot of power, in the form of level N-1 assistants. If they have even a little desire for personal happiness, they might be tempted to trade. Such trade can make the whole bureaucracy corrupt or even harmful.

Sure, a stereotypical "shy frustrated geek" would benefit from moving in the direction of "tough confident warrior". But that's not the only direction and maybe not even the best. How about this one: "speaks ten languages, talks fast, loves to haggle?" The world is so wide! A nice writing exercise i...(read more)

I'm firmly on the side of guess culture on this. If you're good enough at reading my nonverbal signals that you can punch me without making me feel defensive or creeped out, go for it. If you aren't good enough, you aren't good enough, and I won't hesitate to tell you as much.

Our rules don't draw that boundary at the moment, and I'm not even sure how it could be phrased. Do you have any suggestions?

The prize isn't affiliated with MIRI or any other organization, it's just us. But maybe we can figure something out. Can you email me (vladimir.slepnev at gmail)?

Accepted, gave some feedback in the comments