a thing i don’t get is what is new. i mean, computers have long been much, much “smarter” than humans in, for example, their ability to perform arithmetic, or to remember things. recent AI tools are interesting for sure, but what superior competence of their makes these new systems so threatening, compared to older superior competences?
@interfluidity It's because they're starting to bridge the gap between logic and art/emotion. We expect machines to excel at logic but fail at art (without serious help from humans). But now, whether they're truly creating/feeling or just assembling/simulating, the end result is a product that defies our expectations of what machines are capable of.
@LouisIngenthron they certain have astonished us! during the industrial revolution, though, i wonder whether there weren’t analogous moments of just awe that machines could do these things that previously only humans — and it took so many humans! — could do.
@LouisIngenthron the emotion question is interesting. so far, the LLM producers have really tried to limit apparent expression of emotion. (ChatGPT has become very neutral over the months, obviously Microsoft’s “Sidney” was clipped post NYT interview.) but even when they are restricted to HR-speak, we are somewhat awestruck. and it will be interesting to see how we’ll react to apps? bots? services? that aren’t tied back from reproducing (producing?) emotion.
@interfluidity The text-based AIs are held back in that way, yes, but consider the image and music generating AIs.
@LouisIngenthron i guess i feel like we attribute a capacity to experience and therefore experience emotion more to the LLMs than the art-bots. the image generators produce emotionally evocative things for sure, but i don’t find myself anthropomorphizing them, it’s easier to imagine a surprising but mere statistical blender to whose outputs i attribute emotion than it is when i am actually “talking” to the blender.