on what we should spend our efforts
a little while back, I got into a conversation with a friend, Alex, about research. we were both sort of bemoaning the directions that AI is pursuing/has pursued. I was making the argument that building a general learning machine, or general human-like intelligence, or whatever you want to call the grand vision for AI, isn't necessarily the best use of our time. not only does it seem (at least currently) like an untenable goal, but I'm not entirely convinced that creating such an intelligent machine would be of use. Alex then asked on what research efforts we should be spending our human, monetary, and temporal resources. after a bit of consideration, I came up with three things. what follows is roughly a transcript of the chat log from our conversation. I've added a few things, but this is largely how my side of the conversation unfolded.
ok, I think there are three main areas where I'd like to see work done
1) the mind-body problem, reconciling objectivist/subjectivist myths, and breaking out of Cartesian dualism
2) interdisciplinarity and epistemological pluralism
3) sustainable design, environmental psychology, and reducing/reversing global climate change
1) is something that's sort of plagued science for hundreds of years - how does our subjective experience of consciousness arise from the physical reality of our brains. we've done lots of brain mapping, but we're still pretty far off from understanding how what we experience as daily life is linked to our physical world. I might posit that the physical world can only explain so much, and that seeing a strong Cartesian split between the mind and the body (me in here vs the world out there that I'm studying) as science tends to do might be harmful in gaining a better understanding of the world. we should look for ways to advance our understanding that might not fit into the traditional scientific paradigm. this is sort of a segue into 2).
2) has to do with advancing knowledge. any theory is a lens; some theories put certain aspects of a situation into focus while they blur or entirely occlude others. rather than trying to come up with a single theory or model for the entire world (quantum-gravity in physics (or string theory or whatever), a theory of complex systems, etc.), we should look for ways that different theoretical approaches can complement one another. all models all wrong, some are just more wrong than others (that's a quote from some George Box, a statistician).
3) is really, to some extent, the moral imperative of our time. there absolutely no doubt that we are very rapidly nearly destroying the livability of this planet. if we want a home for our children and the future of our species, we absolutely need to put as much effort as possible into salvaging the environment. this is both from a technical standpoint and from a social standpoint. video cameras are smaller, lighter, faster, have more battery life, take higher quality pictures, and are cheaper than they were even 2 or 3 years ago, yet we don't see this level of innovation with cars. other countries have improved fuel efficiency, but in the US... let's just say that the model-T got 28 mpg. we can do better. cars are a convenient example, but it's not just cars, it's alternative energy in general. but it's not just a technological solution for the problem. society has to support this and, as a whole, think it's really important. I bike almost everywhere that I can; I choose dr's offices and stores that are within biking distance; I live close to campus. my girlfriend bikes to work at least 3 days a week. if everyone made small changes like this, it would have almost as much an effect, if not more so, than government regulation. if every person changed one bulb in their house from traditional incandescent to energy efficient, it would reduce annual emissions by something like 10%, and that's just one bulb. understanding individual and social motivations to engender these sorts of changes will, I believe, be crucial in preserving the environment.
after that, we got into a discussion about how the general thinking machine of AI's grand vision might be a useful, integral, or possible necessary part of achieving parts of each of those goals. however, that unfortunately got cut short and we never resumed the conversation. however, I thought that archiving the above here would be sort of a good way to capture/archive what I think is important right now, as well as hopefully generate some discussion with those who might (dis)agree.
ok, I think there are three main areas where I'd like to see work done
1) the mind-body problem, reconciling objectivist/subjectivist myths, and breaking out of Cartesian dualism
2) interdisciplinarity and epistemological pluralism
3) sustainable design, environmental psychology, and reducing/reversing global climate change
1) is something that's sort of plagued science for hundreds of years - how does our subjective experience of consciousness arise from the physical reality of our brains. we've done lots of brain mapping, but we're still pretty far off from understanding how what we experience as daily life is linked to our physical world. I might posit that the physical world can only explain so much, and that seeing a strong Cartesian split between the mind and the body (me in here vs the world out there that I'm studying) as science tends to do might be harmful in gaining a better understanding of the world. we should look for ways to advance our understanding that might not fit into the traditional scientific paradigm. this is sort of a segue into 2).
2) has to do with advancing knowledge. any theory is a lens; some theories put certain aspects of a situation into focus while they blur or entirely occlude others. rather than trying to come up with a single theory or model for the entire world (quantum-gravity in physics (or string theory or whatever), a theory of complex systems, etc.), we should look for ways that different theoretical approaches can complement one another. all models all wrong, some are just more wrong than others (that's a quote from some George Box, a statistician).
3) is really, to some extent, the moral imperative of our time. there absolutely no doubt that we are very rapidly nearly destroying the livability of this planet. if we want a home for our children and the future of our species, we absolutely need to put as much effort as possible into salvaging the environment. this is both from a technical standpoint and from a social standpoint. video cameras are smaller, lighter, faster, have more battery life, take higher quality pictures, and are cheaper than they were even 2 or 3 years ago, yet we don't see this level of innovation with cars. other countries have improved fuel efficiency, but in the US... let's just say that the model-T got 28 mpg. we can do better. cars are a convenient example, but it's not just cars, it's alternative energy in general. but it's not just a technological solution for the problem. society has to support this and, as a whole, think it's really important. I bike almost everywhere that I can; I choose dr's offices and stores that are within biking distance; I live close to campus. my girlfriend bikes to work at least 3 days a week. if everyone made small changes like this, it would have almost as much an effect, if not more so, than government regulation. if every person changed one bulb in their house from traditional incandescent to energy efficient, it would reduce annual emissions by something like 10%, and that's just one bulb. understanding individual and social motivations to engender these sorts of changes will, I believe, be crucial in preserving the environment.
after that, we got into a discussion about how the general thinking machine of AI's grand vision might be a useful, integral, or possible necessary part of achieving parts of each of those goals. however, that unfortunately got cut short and we never resumed the conversation. however, I thought that archiving the above here would be sort of a good way to capture/archive what I think is important right now, as well as hopefully generate some discussion with those who might (dis)agree.
Labels: AI, design, environment, interdisciplinary, musings
posted by Jystar at
23:23
0 Comments:
Post a Comment
<< Home