I love how you track the old nativist/rationalist dichotomy onto current AI debates. I am wondering how you think the question of grounding plays out across the historical divide. The old nativists need sense data to ground language. Kant--not so much--or rather, grounding is immanent through the transcendentals. A Kantian would object at my statement here, but this is how I, a cautious realist, choose to interpret Kant.
But then things get even more interesting when we crossover into the new rationalism where there are emergent thought structures--emergent in the sense of emanating out of probabilistic calculations--that do not appear to need any sort of grounding to have communicative purchase.
I am wondering how you work out these issues. I am thinking about writing about connectionism/symbolism and the importance of the debate for educational contexts --- and I would appreciate your insider view on these issues.
Cheers Nick! I think a piece on connectionism and symbolism is a great one from an educational perspective as it gets at a core dividing line in AI research, says something about how people learn, and allows you to draw into focus the (quite blurry) line between those two things. (Though I am a bit biased given I use the constructivist approach in my research pretty often!)
One thing that I didn’t discuss was the relationship between difference and reference, and whether we think meaning can be derived without any referential quality. It seems to me that we might be in a situation in which large models are extracting ‘meaning’ solely through difference (ie the relationship between tokens), which feels quite different to my own model of human understanding. That being said, it’s not all that clear to me ‘real’ understanding is any more sufficient than simulated understanding as far as the capability to process information, practice some sort of reflection or self-reference, or act successfully is concerned.
I hear what you are saying. Scare quotes are coming across loud and clear. We are in this weird structuralist moment in terms of similarity and difference. I too think that it is still a leap to actual information processing and definitely self reference.
Another great post, Harry.
I love how you track the old nativist/rationalist dichotomy onto current AI debates. I am wondering how you think the question of grounding plays out across the historical divide. The old nativists need sense data to ground language. Kant--not so much--or rather, grounding is immanent through the transcendentals. A Kantian would object at my statement here, but this is how I, a cautious realist, choose to interpret Kant.
But then things get even more interesting when we crossover into the new rationalism where there are emergent thought structures--emergent in the sense of emanating out of probabilistic calculations--that do not appear to need any sort of grounding to have communicative purchase.
I am wondering how you work out these issues. I am thinking about writing about connectionism/symbolism and the importance of the debate for educational contexts --- and I would appreciate your insider view on these issues.
Cheers Nick! I think a piece on connectionism and symbolism is a great one from an educational perspective as it gets at a core dividing line in AI research, says something about how people learn, and allows you to draw into focus the (quite blurry) line between those two things. (Though I am a bit biased given I use the constructivist approach in my research pretty often!)
One thing that I didn’t discuss was the relationship between difference and reference, and whether we think meaning can be derived without any referential quality. It seems to me that we might be in a situation in which large models are extracting ‘meaning’ solely through difference (ie the relationship between tokens), which feels quite different to my own model of human understanding. That being said, it’s not all that clear to me ‘real’ understanding is any more sufficient than simulated understanding as far as the capability to process information, practice some sort of reflection or self-reference, or act successfully is concerned.
I hear what you are saying. Scare quotes are coming across loud and clear. We are in this weird structuralist moment in terms of similarity and difference. I too think that it is still a leap to actual information processing and definitely self reference.