All of za3k's Comments + Replies

za3k00

Just to be explicit: The first AI type, which is "time consistent", would probably press the button. The second AI type, which does worse in impossible worlds, and better in world we don't know are impossible, would not press the button.

0Benya
Yes, that's what I meant.
za3k00

My thinking style has changed twice in my lifetime; I'm going to ramble on about this for a while. I will do a better job of describing me now than when I was younger, because I have a bad memory and it's hard to remember how I used to think.

When I was young (up until the age of 10-12) I just thought in thoughts. I was very quick back then; I could do mental arithmetic and problem solving much faster than I can now, for instance.

Then for just a few years, I started thinking visually. I read all the time during this period; I probably read a ten times as... (read more)

za3k40

I think it would be interesting, regardless of whether it's useful. I'd also like to hear about some non-success stories. It's good to know what to avoid or the limits of one's tools as well.

za3k00

I followed this link, and found the blog of one of the "truth wizards" from the study. She writes about the Amanda Knox case. It seems to entirely focus on Amanda Knox.

7komponisto
This has been mentioned before; I'll reiterate my reaction in more detail here. First of all, there is very little "Truth Wizard" analysis of Amanda Knox on that blog (whatever one thinks about the strength of such evidence in the first place). There are several posts about the case, but in only one of them does the author actually attempt to apply her own "lie-detecting" skills to Knox. (In particular, the most recent post on the case just consists of the author's commentary on someone else's argument that Knox is a sociopath; contrary to orthonormal, there is no claim by the author that she herself has detected sociopathy.) The one post where the author does analyze Knox concerns her statement at Guede's trial, of which only audio (not video) is available. (Of Knox's videotaped testimony at her own trial, the author says: "...without hearing the questions asked of Knox, it is impossible to identify if she is lying." -- emphasis added.) Thus, there is no data about facial expression, which is apparently an important component of the author's technique. Hence confidence in this analysis must be presumably be lowered from what it would be if the author were working from a video recording. But in any case, the reasoning in that post is awful. To the extent the author is skilled in detecting lies, she is obviously not particularly skilled in explaining how she arrives at her conclusions. Here is an example: The author seems to be expecting Amanda's memory of an incident to improve over time. Now, I'm not an expert on memory, but this is directly contrary to my understanding of how it works. In fact, (to invoke my own memory here) I distinctly recall Eliezer mentioning once that memories are re-created each time we remember something. If this is true, it implies that memories -- even if they become more vivid! -- would become less entangled with reality over time, not more; which is anyway what you would expect from....physics. Here is another, well, "red flag", co
za3k70

That's an interesting take. She clearly loves me and my siblings and has never hurt anyone to the best of my knowledge, besides. So, it wasn't an uncomfortable topic--only a bit of an odd position to be in.

Although, I also have to point out adoption does not carry the death penalty, so I can imagine a situation in which my hypothetical parent opts not to kill me because they think the fuzz will catch them.

za3k20

That's very unlikely, I think. She's not interested in rationalism.

za3k10

Allow to to start by saying I enjoyed this post and think Yvain makes an interesting point. It may help explain why rumor spreads well. I have however one difference of opinion, which is that if a person adds "I arrived at this belief through evidence", I would believe their statement more. I would assume they are talking about a non-Bayesian, layman's version of "evidence." (If they're talking about Bayesian evidence, they're probably Bayesian and this is also a mark in their favor)

For instance, in mathematics it's common to state... (read more)

za3k80

My mother made this argument to me probably when I was in high school. Given my position as past infanticide candidate, it was an odd conversation. For the record, she was willing to go up to two or six years old, I think.

And let us not forget the Scrubs episode she also agreed with: "Having a baby is like getting a dog that slowly learns to talk."

6MichaelVassar
Sounds like it would be interesting to have your mother make some comments on LW, if you think she would be interested.

My mother made this argument to me probably when I was in high school. Given my position as past infanticide candidate, it was an odd conversation.

Hey, now you know you were kept around because you were actually wanted, not out of a dull sense of obligation. It's like having a biological parent who is totally okay with giving up children for adoption - and stuck around!

wnoise170

I have said before "I'm a moderate on abortion -- I feel it should be okay up to the fifth trimester." While this does shock people into adjusting what boundaries might be considered acceptable, I no longer think it is something useful to say in most fora. Too much chance of offending people and just causing their brains to shut off.