6 ms·

Two AI Pioneers. Two Bizarre Suicides. What Really Happened?

82 points by terpua 19y ago

bayareaguy 19y ago
More articles should use the textbox-in-a-frame scheme like this one did to quote things at length.
- eru 19y ago
  Why not quote them without a frame at length?
  - bayareaguy 19y ago
    I only like that for short things. One of these days I hope to find a way to get a browser to follow a link by opening a smaller window/area directly in the text. That way I don't entirely lose the visual context on the screen. I've been told the original Xanadu system was designed that way. Enso also uses this idea in places. Now that the http://www.humanized.com http://www.humanized.com developers have joined Mozilla, I'm somewhat optimistic that this may actually happen.
- PJCrosier 19y ago
  I really liked the linking used, it's nice to get quotes to put things in perspective but this took you into the story and gave you all these other documents and sites to explore. It was all kind of creepy really, I felt like a voyeur almost.. excellent article.
DarrenStuart 19y ago
truely a very sad story.
rkts 19y ago
This story and the mention of MIT's "high suicide rate" remind me of something I've wondered about. Depression and suicide seem to be common among very smart people. Does anyone know why?
- palish 19y ago
  Because when you're smart enough to realize you're not quite smart enough, it's crushingly depressing.
  - mojuba 19y ago
    Why should that be depressing? It seems to me quite normal that I'm not smart enough for many hard problems - why should that drive me crazy?
    - kirubakaran 19y ago
      You were smart enough to realize that you need not be depressed when you are smart enough to realize that you are not quite smart enough.
- randallsquared 19y ago
  I don't know. However, here's some hypotheses, some combination of which may be most of the reason: Even if depression and suicide attempts aren't more common, very smart people may be enough more competent that they succeed more often. Most attempts at suicide don't actually result in death. It could be that the easy wins in building a smarter brain happen to lead to a genetic predisposition to depression. It could be that smarter people are more easily able to internalize that there's no ultimate point to life, and less able to just have faith.
  - eru 19y ago
    So smarter people might just be better at suicide?
- hello_moto 19y ago
  You can see from their traits: 1) Super smart 2) Actively thinking every single day 3) Thinking how to solve hard problems 4) Too involved/focused with the area field they are in When people called that RMS is nuts, sometime I wonder, is he really that nuts? He's fighting for a better world of software. But at the same time, he doesn't fight over super-hard-algorithm. RMS is still in touch with humans albeit he's fighting/arguing with them CEOs/PRs. The rate of suicide is high among Grad/PhD students because they have to "think outside the box" every single day. They have to find a "breakthrough". Sometime, "thinking outside the box" means to break all the logics and common sense around you. When you start bending logics, there's a good chance the situation will change the way you perceive things. You began to accept your "new logic" (as in this case, human is just like a program, disease equals to bug).
Alex3917 19y ago
If you want to know what mindpixel actually spent his time thinking about and how others treated him, this k5 article is fairly representative: http://www.kuro5hin.org/story/2005/6/11/224459/073 http://www.kuro5hin.org/story/2005/6/11/224459/073
- boredguy8 19y ago
  Beat me to the k5 connection. But Mindpixel consistently demonstrated an inability to translate his claims into anything remotely resembling reasonable thought. There's a now-lost-forever post in which he claims to have predicted a terrorist attack. http://www.kuro5hin.org/comments/2006/8/21/121745/650?pid=128#135 http://www.kuro5hin.org/comments/2006/8/21/121745/650?pid=12... certainly sums up the silliness.
  - Alex3917 19y ago
    I think the general consensus was that he was a really intelligent and creative guy, but he lacked the discipline to sit down and spend 20 years proving his theories. I remember the post you are talking about though. Supposedly he used a Bayesian filter on real time Google searches to predict two gas station bombings in Baghdad. It's a shame that story didn't get front paged, it was kind of interesting even if it was made up. It's really weird though going back and reading my conversations with him on K5. I'm sure a lot of people here probably knew him though; it would be almost impossible not to have had a conversation with him if you were a regular social news user circa 2000. The guy posted constantly on Slashdot, K5, Joel on Software, Usenet, etc.
    - boredguy8 19y ago
      That's not at all accurate. He was looking at queries that hit GAC and /post facto/ saw a correlation. Not exactly brilliance, imho.
      - Alex3917 19y ago
        If true, this Slashdot comment is also telling: http://slashdot.org/comments.pl?sid=424396&cid=22117930 http://slashdot.org/comments.pl?sid=424396&cid=22117930
doubleplus 19y ago
heh: "How do you keep garbage out without any form of validation mechanism? ... All you have to do is try to [imagine] Slashdot without the moderation system to see what's going to happen to your database."
icky 19y ago
'M-x doctor' made them do it... ;-)
- eru 19y ago
  My Emacs doctor cares: > I am the psychotherapist. Please, describe your problems. Each time you are finished talking, type RET twice. I want to commit suicide. > If you are really suicidal, you might want to contact the Samaritans via E-mail: jo@samaritans.org or, at your option, anonymous E-mail: samaritans@anon.twwells.com . Or find a Befrienders crisis center at http://www.befrienders.org/ http://www.befrienders.org/ . I would appreciate it if you would continue.
tlrobinson 19y ago
Antitrust, anyone? (pretty bad movie, I know, but it makes you wonder...)
- dfranke 19y ago
  Why do so many people rag on that movie? Are they expecting it to be some kind of epic manifesto of open source culture, and then getting let down? It's just entertainment with a few extra morsels for hackers. When you look at it that way, it's easily in the top 20% of what comes out of Hollywood.
rms 19y ago
Existential depression is something we all deal with just fine. It doesn't cause suicide. In this case, presumably unmedicated serious bipolar disorder caused suicide. There is a big difference and a life probably would have been saved if someone would have just taken their lithium.
bayareaguy 19y ago
The story mentioned CyC. I haven't heard about that in at least a decade. Wikipedia says OpenCyc 1.0 was released in July 2006. Google turned up one recent article: - http://newmobilecomputing.com/story/17331/Interview_with_the_OpenCyc_Guys http://newmobilecomputing.com/story/17331/Interview_with_the... but that doesn't say much about any practical applications. There also seems to be a web service for it here: - http://www.cycfoundation.org/blog/?page_id=13 http://www.cycfoundation.org/blog/?page_id=13 I wonder if anyone has found a good application for it. The Texai project sounds like it is trying to do the same thing these people were doing using text from Wikipedia and terms from OpenCyC. - http://www.texai.org/blog/about/texai-project http://www.texai.org/blog/about/texai-project
ojbyrne 19y ago
I met Chris McKinstry in Winnipeg, Canada in 1995. I was not very impressed - he wanted to start a course on linux sysadmin or something... - with me doing pretty well everything except collecting the money... and he couldn't even organize it properly. He struck me as a scammer, though obviously it's only a thin line between that and "Internet entrepreneur."
- marvin 19y ago
  ...and an even thinner line between scammer and AI guru. At least in web startups your popularity will decide whether you are doing something worthwhile or not. In academia, promises that "it will work tomorrow, I swear!" has historically worked very well while providing few results.

82 points by terpua 19y ago

bayareaguy 19y ago

More articles should use the textbox-in-a-frame scheme like this one did to quote things at length.

eru 19y ago
Why not quote them without a frame at length?
- bayareaguy 19y ago
  I only like that for short things. One of these days I hope to find a way to get a browser to follow a link by opening a smaller window/area directly in the text. That way I don't entirely lose the visual context on the screen. I've been told the original Xanadu system was designed that way. Enso also uses this idea in places. Now that the http://www.humanized.com http://www.humanized.com developers have joined Mozilla, I'm somewhat optimistic that this may actually happen.
PJCrosier 19y ago
I really liked the linking used, it's nice to get quotes to put things in perspective but this took you into the story and gave you all these other documents and sites to explore. It was all kind of creepy really, I felt like a voyeur almost.. excellent article.

DarrenStuart 19y ago

truely a very sad story.

rkts 19y ago

This story and the mention of MIT's "high suicide rate" remind me of something I've wondered about. Depression and suicide seem to be common among very smart people. Does anyone know why?

palish 19y ago
Because when you're smart enough to realize you're not quite smart enough, it's crushingly depressing.
- mojuba 19y ago
  Why should that be depressing? It seems to me quite normal that I'm not smart enough for many hard problems - why should that drive me crazy?
  - kirubakaran 19y ago
    You were smart enough to realize that you need not be depressed when you are smart enough to realize that you are not quite smart enough.
randallsquared 19y ago
I don't know. However, here's some hypotheses, some combination of which may be most of the reason: Even if depression and suicide attempts aren't more common, very smart people may be enough more competent that they succeed more often. Most attempts at suicide don't actually result in death. It could be that the easy wins in building a smarter brain happen to lead to a genetic predisposition to depression. It could be that smarter people are more easily able to internalize that there's no ultimate point to life, and less able to just have faith.
- eru 19y ago
  So smarter people might just be better at suicide?
hello_moto 19y ago
You can see from their traits: 1) Super smart 2) Actively thinking every single day 3) Thinking how to solve hard problems 4) Too involved/focused with the area field they are in When people called that RMS is nuts, sometime I wonder, is he really that nuts? He's fighting for a better world of software. But at the same time, he doesn't fight over super-hard-algorithm. RMS is still in touch with humans albeit he's fighting/arguing with them CEOs/PRs. The rate of suicide is high among Grad/PhD students because they have to "think outside the box" every single day. They have to find a "breakthrough". Sometime, "thinking outside the box" means to break all the logics and common sense around you. When you start bending logics, there's a good chance the situation will change the way you perceive things. You began to accept your "new logic" (as in this case, human is just like a program, disease equals to bug).

Alex3917 19y ago

If you want to know what mindpixel actually spent his time thinking about and how others treated him, this k5 article is fairly representative: http://www.kuro5hin.org/story/2005/6/11/224459/073 http://www.kuro5hin.org/story/2005/6/11/224459/073

boredguy8 19y ago
Beat me to the k5 connection. But Mindpixel consistently demonstrated an inability to translate his claims into anything remotely resembling reasonable thought. There's a now-lost-forever post in which he claims to have predicted a terrorist attack. http://www.kuro5hin.org/comments/2006/8/21/121745/650?pid=128#135 http://www.kuro5hin.org/comments/2006/8/21/121745/650?pid=12... certainly sums up the silliness.
- Alex3917 19y ago
  I think the general consensus was that he was a really intelligent and creative guy, but he lacked the discipline to sit down and spend 20 years proving his theories. I remember the post you are talking about though. Supposedly he used a Bayesian filter on real time Google searches to predict two gas station bombings in Baghdad. It's a shame that story didn't get front paged, it was kind of interesting even if it was made up. It's really weird though going back and reading my conversations with him on K5. I'm sure a lot of people here probably knew him though; it would be almost impossible not to have had a conversation with him if you were a regular social news user circa 2000. The guy posted constantly on Slashdot, K5, Joel on Software, Usenet, etc.
  - boredguy8 19y ago
    That's not at all accurate. He was looking at queries that hit GAC and /post facto/ saw a correlation. Not exactly brilliance, imho.
    - Alex3917 19y ago
      If true, this Slashdot comment is also telling: http://slashdot.org/comments.pl?sid=424396&cid=22117930 http://slashdot.org/comments.pl?sid=424396&cid=22117930

doubleplus 19y ago

heh: "How do you keep garbage out without any form of validation mechanism? ... All you have to do is try to [imagine] Slashdot without the moderation system to see what's going to happen to your database."

icky 19y ago

'M-x doctor' made them do it... ;-)

eru 19y ago
My Emacs doctor cares: > I am the psychotherapist. Please, describe your problems. Each time you are finished talking, type RET twice. I want to commit suicide. > If you are really suicidal, you might want to contact the Samaritans via E-mail: jo@samaritans.org or, at your option, anonymous E-mail: samaritans@anon.twwells.com . Or find a Befrienders crisis center at http://www.befrienders.org/ http://www.befrienders.org/ . I would appreciate it if you would continue.

tlrobinson 19y ago

Antitrust, anyone? (pretty bad movie, I know, but it makes you wonder...)

dfranke 19y ago
Why do so many people rag on that movie? Are they expecting it to be some kind of epic manifesto of open source culture, and then getting let down? It's just entertainment with a few extra morsels for hackers. When you look at it that way, it's easily in the top 20% of what comes out of Hollywood.

rms 19y ago

Existential depression is something we all deal with just fine. It doesn't cause suicide. In this case, presumably unmedicated serious bipolar disorder caused suicide. There is a big difference and a life probably would have been saved if someone would have just taken their lithium.

bayareaguy 19y ago

The story mentioned CyC. I haven't heard about that in at least a decade. Wikipedia says OpenCyc 1.0 was released in July 2006. Google turned up one recent article: - http://newmobilecomputing.com/story/17331/Interview_with_the_OpenCyc_Guys http://newmobilecomputing.com/story/17331/Interview_with_the... but that doesn't say much about any practical applications. There also seems to be a web service for it here: - http://www.cycfoundation.org/blog/?page_id=13 http://www.cycfoundation.org/blog/?page_id=13 I wonder if anyone has found a good application for it. The Texai project sounds like it is trying to do the same thing these people were doing using text from Wikipedia and terms from OpenCyC. - http://www.texai.org/blog/about/texai-project http://www.texai.org/blog/about/texai-project

ojbyrne 19y ago

I met Chris McKinstry in Winnipeg, Canada in 1995. I was not very impressed - he wanted to start a course on linux sysadmin or something... - with me doing pretty well everything except collecting the money... and he couldn't even organize it properly. He struck me as a scammer, though obviously it's only a thin line between that and "Internet entrepreneur."

marvin 19y ago
...and an even thinner line between scammer and AI guru. At least in web startups your popularity will decide whether you are doing something worthwhile or not. In academia, promises that "it will work tomorrow, I swear!" has historically worked very well while providing few results.