GreenRoot

Karma: 408

GreenRoot 10 Dec 2009 2:09 UTC
4 points
on: You Be the Jury: Survey on a Current Event
1. Knox guilt: 80%
2. Sollecito guilt: 80%
3. Guede guilt: 99%
4. prob of agreement: 0.8
5. I was unfamiliar with the case before hearing about the verdicts on the news. I spent about 10 minutes on the Wikipedia article, then20 minutes on each of the sites you linked, and made these judgments before reading others’ comments.
Guede’s guilt seems almost beyond doubt, given the DNA evidence, his implausible story, and his flight without calling police. Regarding Knox & Sollecito, I’m convinced of the prosecution’s version of events mainly by the cell-phone record evidence and the many conflicting statements given by Knox and Sollecito over time, with variation that goes beyond my (only moderately informed) expectations for bad memory or coercive interrogation.

Meta-evidence judgments: in the pro-guilt site you linked, it took a while to find concrete info, but when I found the Prosecution’s case->Facts Presented section, I found a lot of very specific detail regarding the exact times cell phones were turned on and off and the times and durations of calls made, as well as detail regarding the forensic evidence and the credentials of the experts who testified about it. The pro-innocent site lacked such specifics and didn’t seem to account for all the damning details. Both sites seemed very distorted by affection for Kercher (pro-guilt) and Knox (pro-innocence) and their nationalities.

I read the pro-guilt site before the pro-innocent site, and I noticed this causing bias: as I read the pro-innocence site, I felt myself internally “rooting” against them, expecting that they couldn’t answer all the evidence and (maybe not surprising) finding that this was so. I tried to adjust for this in my judgments.

I have a low prior for this mix of murderers committing the crime together, and a high prior for a guilty verdict given by a jury, and feel like these two facts approximately cancel out.

After reading others’ comments, I was surprised that most others leaned toward Knox’s and Sollecito’s innocence. I would update my judgments only marginally because of this, because I read the linked sites more closely than others’ claimed to. I do however revise my estimate of agreement with komponisto down a bit to 0.6.

p.s. This is a very interesting exercise, which led me to register and comment for the first time after lurking for several months. I am very curious about your (komponisto’s) judgment and reasoning.

GreenRoot 10 Dec 2009 19:40 UTC
0 points
in reply to: Jack’s comment on: You Be the Jury: Survey on a Current Event

You weren’t troubled by the lack of motive...

Hmm, thinking back, I didn’t consider motive too closely. I rather got caught up in the evidence of Amanda’s evasion. In retrospect, this would lower my probabilities.

...and the fact that after covering up her involvement in a crime but intentionally leaving evidence of Guede’s presence Knox went on to point the finger at someone who couldn’t have been involved after being coerced by the police?

I didn’t judge that evidence against Guede was left intentionally; I could easily have missed details indicating this. Knox’s fingering of Lumumba seemed to me to be a natural part of a collection of inconsistent attempts to establish an alibi and deflect suspicion.

GreenRoot 10 Dec 2009 19:59 UTC
6 points
in reply to: GreenRoot’s comment on: You Be the Jury: Survey on a Current Event
After reading more comments, I’ve updated my probabilities significantly. Here’s what influences me:
1. Different commenters focus on different subsets of evidence. This makes me suspect that my own focal subset was incidental and probably depended on what order I encountered various claims.
2. Many items of evidence presented as fact (and which I relied on), such as cleaning supply shopping and the contents of the washing machine, are said by others to be rumors and were never presented at trial. This undermined a lot of what I based my judgment on.
3. The judgment of people who have followed the case very closely (e.g. jenmarle) is that Knox and Sollecito are innocent.
Together, these things make me throw my hands up in the air. I don’t think I can clear things up without spending much more time on it, definitely not without seeking out new sources, and I don’t know where to look. I now believe: Knox: 50%, Sollecito: 50%, Guede: 99%.
What links here?
- Amanda Knox: post mortem by gwern (20 Oct 2011 16:10 UTC; 48 points)

GreenRoot 10 Dec 2009 20:05 UTC
1 point
on: You Be the Jury: Survey on a Current Event
For those of you who were not familiar with the case, which side’s arguments did you read first?

Did it feel like this order influenced your reasoning or final judgment?

(I read pro-guilt ideas first, and feel now that this ordering contributed to my judgment that Knox and Sollecito are guilty. I wonder how conviction rates would change in the US if we let the defense go first, i.e. what are the relative influence values of the right to speak first vs. last?)

GreenRoot 4 Feb 2010 19:02 UTC
3 points
on: Open Thread: February 2010
How about per-capita post scoring?

Why not divide a post’s number of up-votes by the number of unique logged-in people who have viewed it? This would correct for the distortion of scores caused by varying numbers of readers. Some old stuff is very good but not read much, and scores are in general inflating as the Less Wrong population grows.

I think such a change would be orthogonal to karma accounting; I’m only suggesting a change in the number displayed next to each post.

GreenRoot 5 Feb 2010 16:19 UTC
6 points
on: Debate tools: an experience report
I looked into this sort of software a couple years ago as part of a information visualization seminar at Stanford. People have been creating research prototypes and betas for a long time, aiming to support everything from education to the formalization of legal dispute resolution (“computational law”). I wasn’t impressed with what I saw in my review then, and looking around again today, it doesn’t look like things have changed much. I have never seen anything quantitative.

You can find descriptions of research systems and small-scale studies in the education vis and info-vis literatures. (InfoVis, VAST, etc.)

A book: http://www.visualizingargumentation.info/ A new journal: http://www.tandf.co.uk/journals/tarc

GreenRoot 19 Feb 2010 0:17 UTC
4 points
on: Open Thread: February 2010, part 2
What do you have to protect?

Eliezer has stated that rationality should not be end in itself, and that to get good at it, one should be motivated by something more important. For those of you who agree with Eliezer on this, I would like to know: What is your reason? What do you have to protect?

This is a rather personal question, I know, but I’m very curious. What problem are you trying to solve or goal are you trying to reach that makes reading this blog and participating in its discourse worthwhile to you?

GreenRoot 19 Feb 2010 1:20 UTC
2 points
in reply to: h-H’s comment on: Open Thread: February 2010, part 2
Yes, this is what I was curious about, thanks. I’ve seen others cite humanity’s existential risks as their motivations too (mostly uFAI, not as much nuclear war or super-flu or meteors). I’m like you in that for me it’s definitely a mix of learning and entertainment.

GreenRoot 22 Feb 2010 21:45 UTC
5 points
on: Akrasia Tactics Review
Some things I’ve tried:
1. Private Resolutions: −4 Going off by myself, talking through the seriousness of the situation, and promising myself to do better / work harder. Usually has brief positive impact, but eventually I backslide. The next time around, the resolution has to be even more emphatic, because I know the previous level of seriousness wasn’t enough. This is a ratchet of self-blame that I really can’t take anymore.
2. LeechBlock: short term +2, long term +0 I tried this for about six months. It helped for most of that time. Toward the end, I just ended up using another browser.
3. Getting Things Done: +4 This one has been a big and lasting help, especially in areas of my life outside of schoolwork. Where I don’t have complicated psychology of fear and habit driving procrastination, I’m limited mainly by keeping track of things. The GTD system has worked very well for me on that. Biggest benefit comes from forcing me to decide what I actually intend to do.
4. Mild accountability: +2 My thesis work is very individual, but I can give high-level reports on how things are going to friends and family, folks I’m not willing to lie to. The desire to give positive reports keeps things moving, but hasn’t been a big drive.
5. Expose myself to hard-working friends: +5 When I see my friends working hard for any reason (against their will or out of internal motivation), it puts my procrastination in perspective and makes it seem silly to me. This helps me just get over it and get work done.
6. Getting enough sleep: +2 I learned years ago that sleep is necessary for willpower. It’s not sufficient for productivity, but if I’m tired, it’s very easy to give in to distraction.
Thanks for the work on this. I look forward to reviewing both the list of techniques and people’s aggregated experiences.
What links here?
- Akrasia Tactics Review by orthonormal (21 Feb 2010 4:25 UTC; 91 points)

GreenRoot 22 Feb 2010 22:45 UTC
12 points
in reply to: GreenRoot’s comment on: Akrasia Tactics Review
One more I didn’t include but have seen others mention:
1. collaboration: +8
Whenever I’m working on a project with other people, especially when we’re in the same place, accountability and shared excitement totally short-circuit procrastination. I’ve only ever procrastinated on solo projects.

GreenRoot 26 Feb 2010 18:36 UTC
2 points
in reply to: Kevin’s comment on: New Year’s Predictions Thread
The term “terrorism” is usually taken to mean an attack on civilians, though as a legal matter, this is far from settled. This definition would exclude the Fort Hood shooting, where the targets were soldiers. In any case, the bet is over non-state, politically motivated killing, which is broader and would include Fort Hood, I think.

GreenRoot 4 Mar 2010 17:45 UTC
10 points
in reply to: Morendil’s comment on: Priors and Surprise
For others’ reference, PT:TLOS = Probability Theory: The Logic of Science, by E. T. Jaynes.

GreenRoot 4 Mar 2010 18:02 UTC
2 points
in reply to: CronoDAS’s comment on: Open Thread: March 2010
Very good. I see this forcing more careful thought by the poster, either now or later, and more skepticism in the blog’s audience.

I’d recommend restating all the terms of the bet in a single comment or another web page, which both of you explicitly accept. This will make things easier to reference eight months from now. Might also be good to name a simple procedure like a poll on the blog to resolve any disagreements (like the definition of “Healthcare reform passes”).

And please, reply again here or make a new open thread comment once this gets resolved. I’d love to hear how it turned out and what the impact on poster’s or other’s beliefs was.

GreenRoot 6 Mar 2010 12:27 UTC
4 points
on: The strongest status signals
It seems to me that in an analysis like this, some distinction should be drawn between people you expect to interact with in the future (people who’s status matters) and strangers. People ignore others in cities all the time, and I think the I’ll-never-see-you-again hypothesis explains it more simply than the you-have-no-power-over-me hypothesis.

GreenRoot 17 Mar 2010 3:54 UTC
29 points
on: Living Luminously

I have already written all of the posts in this sequence, although I may make edits to later ones in response to feedback on earlier ones, and it’s not impossible that someone will ask me something that seems to indicate I should write an additional post.

This preparation sounds great. Thank you for taking such care with the writing, and with providing this introduction. The idea of thorough, regulated introspection is new to me, and I’m looking forward to hearing from somebody who’s put a lot of thought into it.

A site where people (1) do deep original thinking, then (2) spend considerable time and effort to write accessibly about it, and (3) refine the ideas through civil discussion: all of these things are so rare that the combination of them on this site makes it the best philosophy/discussion forum I’ve ever been a part of.

GreenRoot 17 Mar 2010 4:15 UTC
20 points
on: You Are Likely To Be Eaten By A Grue
tl;dr: Why study yourself? Because
1. You’re interesting because you’re complicated and you’re you.
2. You can convincingly ask people to adjust how they deal with you to your peculiarities.
3. You can learn how to improve yourself and thus accomplish your goals.
Fair summary?

Would you say that improved self awareness can help you accomplish your goals directly, without the self-modification discussed in reason #3? Something like: learn more about what makes you happy, then do that stuff. This seems like a bigger payoff than #2 to me, worthy of inclusion in any sales pitch for self-study.

GreenRoot 26 Mar 2010 17:22 UTC
4 points
on: The Shabbos goy
This is a nice and unexpected parallel. I think I may notice more instances now that I know the general pattern. Thanks for writing it up.

Free copy of Feynman’s autobiography for best corny rationalist joke

GreenRoot4 Apr 2010 0:32 UTC

19 points

54 comments1 min readLW link

GreenRoot 4 May 2010 18:22 UTC
0 points
on: Jinnetic Engineering, by Richard Stallman
Nice link, but would be better as on open thread comment.

GreenRoot 7 May 2010 18:39 UTC
2 points
in reply to: Tehom’s comment on: But Somebody Would Have Noticed
I think this point about shorthand arguments and their expansion on demand is very helpful. I’d love to see a top-level post on it, with one or two additional examples.

GreenRoot

Free copy of Feyn­man’s au­to­bi­og­ra­phy for best corny ra­tio­nal­ist joke

Free copy of Feynman’s autobiography for best corny rationalist joke