5 ms·

Making AI Philosophical Again: On Philip E. Agre’s Legacy (2014)

70 points by nkurz 7y ago

nkurz 7y ago
Since there is no other discussion, I'll try to start. I posted this today because someone else has posted one of Phil's 2002 essays: https://news.ycombinator.com/item?id=21808425 https://news.ycombinator.com/item?id=21808425. Phil was a professor at UCLA, and wrote early and presciently about the social aspects of the internet, and how it was going to change society. Then eventually he went silent, and in 2009 he was reported by a relative as "missing". He was reported as "found" a few months later, but in a way that left a lot of mystery as to what had actually happened. I have no personal knowledge, but my impression from the outside is that his work made him so depressed that he gave up on society and just dropped out. I'm sure there were other psychological factors involved, but also at least some elements of a cautionary tale. I learned about Phil from his Red Rock Eater News Service (https://pages.gseis.ucla.edu/faculty/agre/rre.html https://pages.gseis.ucla.edu/faculty/agre/rre.html) which I subscribed to in the mid to late 90's. RRE was a curated collection of links, with commentary and occasionally essays. While I don't know of any direct link, I think of RRE as being one of the philosophical ancestors of Hacker News. Thus since you are reading this on HN, whether you've heard of him or not, you might in some small way owe a small thanks to Phil.
- extra88 7y ago
  RRE was great. I wasn't an aspiring academic but his writings offering advice to grad students looking ahead to an academic career were very interesting.
- schoen 7y ago
  I used to read RRE as an adolescent and would occasionally write replies to Phil about items he included. My replies were not very sophisticated or well-informed; I remember that Phil would always reply and would never sugar-coat his opinion of what I had written. One amazing thing I learned later is about the Buddhist philosopher David Chapman's collaboration with Phil when both of them were researchers at CSAIL. See, e.g., https://meaningness.com/metablog/ken-wilber-boomeritis-artificial-intelligence https://meaningness.com/metablog/ken-wilber-boomeritis-artif...
- lisper 7y ago
  I met Phil when I was a grad student around 1990. (Holy shit, that was thirty years ago!) His work on Pengi [1] was a strong influence on me in no small measure because it just seemed so obviously the Right Thing compared to everything else that was going on at the time. I was mystified that no one had proposed indexical representations sooner, and that the idea wasn't better received. The thing I remember most about Phil was how he introduced himself when we met: "Hi, I'm Phil Agre. Who are you?" Direct, to the point, and to a fellow aspergery person like me, just obviously the Right Thing. Since then I've learned a lot about the world, and at times had my own struggles with existential despair. It makes me very sad to think that Phil dropped out because wasn't able to figure out a better way to navigate reality. [1] https://www.aaai.org/Papers/AAAI/1987/AAAI87-048.pdf https://www.aaai.org/Papers/AAAI/1987/AAAI87-048.pdf
rootbear 7y ago
I knew Phil slightly when I was at the University of Maryland. I think we had a physics class together. I remember him as being quite smart. I hope he is well and happy.
NelsonMinar 7y ago
I miss Phil.
Aaronstotle 7y ago
Not directly related, but as a philosophy undergrad at Berkeley, I was disappointed with the lack of classes/studies dedicated to the field I call: Philosophy of Technology.
nathias 7y ago
I can relate. Has anyone read his works and has an idea on where to start?
tus88 7y ago
We will know AI when a machine blurts out "I think, therefor I am" without any prior input.
- a-priori 7y ago
  Would you expect a human to say that 'without any prior input'?
  - tus88 7y ago
    Eventually, yes. Although that is an interesting philosophical question in itself.
    - thundergolfer 7y ago
      Eventually being when? Billions of humans lived and died without ever saying it, and they had plenty of related “prior input”.
      - tus88 7y ago
        Maybe they had too much junk input to have deep philosophical thoughts :D
jonjacky 7y ago
Phil Agre was a prolific writer with many high-quality contributions on a variety of topics. Here are some samples: How to help someone use a computer: https://pages.gseis.ucla.edu/faculty/agre/how-to-help.html https://pages.gseis.ucla.edu/faculty/agre/how-to-help.html Rationalizations for bad design, a posting to RISKS digest: http://catless.ncl.ac.uk/Risks/7.09.html#subj1 http://catless.ncl.ac.uk/Risks/7.09.html#subj1 Layering, from a course on Information Systems and Design: https://pages.gseis.ucla.edu/faculty/agre/240/week5.html https://pages.gseis.ucla.edu/faculty/agre/240/week5.html Toward a Critical Technical Practice: Lessons Learned in Trying to Reform AI: https://pages.gseis.ucla.edu/faculty/agre/critical.html https://pages.gseis.ucla.edu/faculty/agre/critical.html Notes and Recommendations (from RRE Digest): https://pages.gseis.ucla.edu/faculty/agre/notes.html https://pages.gseis.ucla.edu/faculty/agre/notes.html Red Rock Eater Digest, 1994 -- 2004: http://web.archive.org/web/20040602193512/commons.somewhere.com/rre/ http://web.archive.org/web/20040602193512/commons.somewhere.... The Network Observer, 1994 -- 1996: https://pages.gseis.ucla.edu/faculty/agre/tno.html https://pages.gseis.ucla.edu/faculty/agre/tno.html He wrote a book, Computation and Human Experience, here are some extracts and a chapter summary: https://pages.gseis.ucla.edu/faculty/agre/che-intro.html https://pages.gseis.ucla.edu/faculty/agre/che-intro.html I miss him.
gumby 7y ago
He came to visit my wife and me in Paris a few years before he left the grid. He was already sort of drifting away. Still we had a really fun visit.

70 points by nkurz 7y ago

Since there is no other discussion, I'll try to start. I posted this today because someone else has posted one of Phil's 2002 essays: https://news.ycombinator.com/item?id=21808425 https://news.ycombinator.com/item?id=21808425. Phil was a professor at UCLA, and wrote early and presciently about the social aspects of the internet, and how it was going to change society. Then eventually he went silent, and in 2009 he was reported by a relative as "missing". He was reported as "found" a few months later, but in a way that left a lot of mystery as to what had actually happened. I have no personal knowledge, but my impression from the outside is that his work made him so depressed that he gave up on society and just dropped out. I'm sure there were other psychological factors involved, but also at least some elements of a cautionary tale. I learned about Phil from his Red Rock Eater News Service (https://pages.gseis.ucla.edu/faculty/agre/rre.html https://pages.gseis.ucla.edu/faculty/agre/rre.html) which I subscribed to in the mid to late 90's. RRE was a curated collection of links, with commentary and occasionally essays. While I don't know of any direct link, I think of RRE as being one of the philosophical ancestors of Hacker News. Thus since you are reading this on HN, whether you've heard of him or not, you might in some small way owe a small thanks to Phil.

extra88 7y ago
RRE was great. I wasn't an aspiring academic but his writings offering advice to grad students looking ahead to an academic career were very interesting.
schoen 7y ago
I used to read RRE as an adolescent and would occasionally write replies to Phil about items he included. My replies were not very sophisticated or well-informed; I remember that Phil would always reply and would never sugar-coat his opinion of what I had written. One amazing thing I learned later is about the Buddhist philosopher David Chapman's collaboration with Phil when both of them were researchers at CSAIL. See, e.g., https://meaningness.com/metablog/ken-wilber-boomeritis-artificial-intelligence https://meaningness.com/metablog/ken-wilber-boomeritis-artif...
lisper 7y ago
I met Phil when I was a grad student around 1990. (Holy shit, that was thirty years ago!) His work on Pengi [1] was a strong influence on me in no small measure because it just seemed so obviously the Right Thing compared to everything else that was going on at the time. I was mystified that no one had proposed indexical representations sooner, and that the idea wasn't better received. The thing I remember most about Phil was how he introduced himself when we met: "Hi, I'm Phil Agre. Who are you?" Direct, to the point, and to a fellow aspergery person like me, just obviously the Right Thing. Since then I've learned a lot about the world, and at times had my own struggles with existential despair. It makes me very sad to think that Phil dropped out because wasn't able to figure out a better way to navigate reality. [1] https://www.aaai.org/Papers/AAAI/1987/AAAI87-048.pdf https://www.aaai.org/Papers/AAAI/1987/AAAI87-048.pdf

rootbear 7y ago

I knew Phil slightly when I was at the University of Maryland. I think we had a physics class together. I remember him as being quite smart. I hope he is well and happy.

NelsonMinar 7y ago

I miss Phil.

Aaronstotle 7y ago

Not directly related, but as a philosophy undergrad at Berkeley, I was disappointed with the lack of classes/studies dedicated to the field I call: Philosophy of Technology.

nathias 7y ago

I can relate. Has anyone read his works and has an idea on where to start?

tus88 7y ago

We will know AI when a machine blurts out "I think, therefor I am" without any prior input.

a-priori 7y ago
Would you expect a human to say that 'without any prior input'?
- tus88 7y ago
  Eventually, yes. Although that is an interesting philosophical question in itself.
  - thundergolfer 7y ago
    Eventually being when? Billions of humans lived and died without ever saying it, and they had plenty of related “prior input”.
    - tus88 7y ago
      Maybe they had too much junk input to have deep philosophical thoughts :D

jonjacky 7y ago

Phil Agre was a prolific writer with many high-quality contributions on a variety of topics. Here are some samples: How to help someone use a computer: https://pages.gseis.ucla.edu/faculty/agre/how-to-help.html https://pages.gseis.ucla.edu/faculty/agre/how-to-help.html Rationalizations for bad design, a posting to RISKS digest: http://catless.ncl.ac.uk/Risks/7.09.html#subj1 http://catless.ncl.ac.uk/Risks/7.09.html#subj1 Layering, from a course on Information Systems and Design: https://pages.gseis.ucla.edu/faculty/agre/240/week5.html https://pages.gseis.ucla.edu/faculty/agre/240/week5.html Toward a Critical Technical Practice: Lessons Learned in Trying to Reform AI: https://pages.gseis.ucla.edu/faculty/agre/critical.html https://pages.gseis.ucla.edu/faculty/agre/critical.html Notes and Recommendations (from RRE Digest): https://pages.gseis.ucla.edu/faculty/agre/notes.html https://pages.gseis.ucla.edu/faculty/agre/notes.html Red Rock Eater Digest, 1994 -- 2004: http://web.archive.org/web/20040602193512/commons.somewhere.com/rre/ http://web.archive.org/web/20040602193512/commons.somewhere.... The Network Observer, 1994 -- 1996: https://pages.gseis.ucla.edu/faculty/agre/tno.html https://pages.gseis.ucla.edu/faculty/agre/tno.html He wrote a book, Computation and Human Experience, here are some extracts and a chapter summary: https://pages.gseis.ucla.edu/faculty/agre/che-intro.html https://pages.gseis.ucla.edu/faculty/agre/che-intro.html I miss him.

gumby 7y ago

He came to visit my wife and me in Paris a few years before he left the grid. He was already sort of drifting away. Still we had a really fun visit.