First, I don’t understand why IIT is still popular, Scott Aaronson showed its fatal shortcomings 10 years ago, as soon as it came out
Well, Scott constructed an example for which the theory gives a highly unintuitive result. This isn’t obviously a fatal critique; you could always argue that a lot of theories give some unintuitive results. It’s also the kind of thing you could maybe fix by tweaking the math,[1] rather than tossing out the entire approach.
I believe Tononi is on record somewhere biting the bullet on that point (i.e., agreeing that Scott’s construction would indeed have high Φ, and that that’s okay). But I don’t know where, and I think I already searched for it a few months ago (probably right after IIT4.0 was dropped) and couldn’t find it.
Second, I do not see any difference between experiencing something and claiming to experience something, outside of intentionally trying to deceive someone.
I think this puts you firmly into Camp #1 (though you saying this proves that, at a minimum, the idea wasn’t communicated as clearly as I’d hoped). Like, the introductory dialogue shows someone failing to communicate the difference, so if this difference isn’t intuitively obvious to you, this would be a Camp #1 characteristic.
And like, since the whole point was that [trying to articulate what exactly it means for experience to exist independently of the report] is extremely difficult and usually doesn’t work, I’m not gonna attempt it here.
I believe Tononi is on record somewhere biting the bullet on that point (i.e., agreeing that Scott’s construction would indeed have high Φ, and that that’s okay). But I don’t know where, and I think I already searched for it a few months ago (probably right after IIT4.0 was dropped) and couldn’t find it.
Thanks!
Well, Scott constructed an example for which the theory gives a highly unintuitive result. This isn’t obviously a fatal critique; you could always argue that a lot of theories give some unintuitive results. It’s also the kind of thing you could maybe fix by tweaking the math,[1] rather than tossing out the entire approach.
I believe Tononi is on record somewhere biting the bullet on that point (i.e., agreeing that Scott’s construction would indeed have high Φ, and that that’s okay). But I don’t know where, and I think I already searched for it a few months ago (probably right after IIT4.0 was dropped) and couldn’t find it.
I think this puts you firmly into Camp #1 (though you saying this proves that, at a minimum, the idea wasn’t communicated as clearly as I’d hoped). Like, the introductory dialogue shows someone failing to communicate the difference, so if this difference isn’t intuitively obvious to you, this would be a Camp #1 characteristic.
And like, since the whole point was that [trying to articulate what exactly it means for experience to exist independently of the report] is extremely difficult and usually doesn’t work, I’m not gonna attempt it here.
Though as mentioned in another comment, I haven’t actually read through the construction—I always just trusted Scott here—so maybe I’m wrong.
There’s a link in SA last post on the topic: https://scottaaronson.blog/?p=1823
Thanks! Sooner or later I would have searched until finding it, now you’ve saved me the time.