An incredible interview. For someone new to the field, this interview offers an excellent overview of the early history of AI. I found an interesting tension at work between Winograd's notion of symmetrical awareness and his older commitments to cybernetics. If we cannot really know "what IT is," how can humans experience something like symmetry even in the face of another human? But I sense I might be overreading a passing notion on his part. As a sort of sidebar: I wonder how technological mediations impact the sense of symmetrical awareness. As human-human contact is deferred, complicated, and reimagined via new technology, how is symmetrical awareness altered, erased, and reconfigured? All of this is a long way of saying that I personally feel that these new LLMs have disrupted the possibility of relying on symmetrical awareness as a reliable ground except perhaps in the most local, in-person environments. And that we can say this without necessarily positing that LLMs exhibit anything reminiscent of human consciousness.
> I personally feel that these new LLMs have disrupted the possibility of relying on symmetrical awareness as a reliable ground except perhaps in the most local, in-person environments.
Super interesting line of thinking here. It does seem more and more important to pay attention to what philosophy people have said about the limitations of our knowledge of others' internal states and our inability to even know that those internal states exist in the first place—the classic p-zombie issue! I think the biggest way I see "symmetrical awareness" being reconfigured—and this has been happening for a long time—is the way in which lots of digital technologies present others to us as static online personas that are combinations of photos, personal descriptions, Tweets / writings of other sorts. There are a lot of cases where these representations don't seem very conducive to thinking of people as actual human beings with their own inner worlds—at least it seems that way to me. I think it's importantly different from e.g. letter writing / pen pals because corresponding with someone with something like letters requires a lot more effort and care. But there's also context involved here, e.g. letter writing today isn't letter writing 20 or 200 years ago.
Great interviews lately. I was critical of "monologue-style" episodes in a past comment, but the podcast now just keeps getting better and better.
Please keep sharing feedback if/when you have any—it helps 🙂
this is one of the best interviews yet! Winograd is not only a CS history icon, he's the most amazing CS teacher and researcher!
Thank you Ruth!!
Oh... this is going to be amazing. Can't wait to listen!!!
Hope you enjoy it!
An incredible interview. For someone new to the field, this interview offers an excellent overview of the early history of AI. I found an interesting tension at work between Winograd's notion of symmetrical awareness and his older commitments to cybernetics. If we cannot really know "what IT is," how can humans experience something like symmetry even in the face of another human? But I sense I might be overreading a passing notion on his part. As a sort of sidebar: I wonder how technological mediations impact the sense of symmetrical awareness. As human-human contact is deferred, complicated, and reimagined via new technology, how is symmetrical awareness altered, erased, and reconfigured? All of this is a long way of saying that I personally feel that these new LLMs have disrupted the possibility of relying on symmetrical awareness as a reliable ground except perhaps in the most local, in-person environments. And that we can say this without necessarily positing that LLMs exhibit anything reminiscent of human consciousness.
I'm glad you liked it!
> I personally feel that these new LLMs have disrupted the possibility of relying on symmetrical awareness as a reliable ground except perhaps in the most local, in-person environments.
Super interesting line of thinking here. It does seem more and more important to pay attention to what philosophy people have said about the limitations of our knowledge of others' internal states and our inability to even know that those internal states exist in the first place—the classic p-zombie issue! I think the biggest way I see "symmetrical awareness" being reconfigured—and this has been happening for a long time—is the way in which lots of digital technologies present others to us as static online personas that are combinations of photos, personal descriptions, Tweets / writings of other sorts. There are a lot of cases where these representations don't seem very conducive to thinking of people as actual human beings with their own inner worlds—at least it seems that way to me. I think it's importantly different from e.g. letter writing / pen pals because corresponding with someone with something like letters requires a lot more effort and care. But there's also context involved here, e.g. letter writing today isn't letter writing 20 or 200 years ago.
Don't most pronounce SHRDLU as sure'-duh-loo?