What if humans’ responses are merely probabilistically consistent with a history of sensory experiences? Would this change the significance of human emotions vs apparent emergent emotional responses from LLMs?
emotions regulate motivation, desire, action, behaviour etc.
to be angry is for your sensory-motor system to be primed for aggression; it's for your cognitive systems to be narrowed and focused on analysing high-threat parts of your environment; it is for your memory-formulation to be modulated towards threat recollection etc.
Sure, if an LLM's prompt "be angry" causes it to adopt a threat stance to its environment, to regulate it's theory-of-mind to engage with possible hostile entities, and so on --- then yes, when LLMs are there, I shall concede the point
However, how terrible it would be to start with an analysis of emotions in terms of the capacities of LLMs -- right?
Since if you did that you'd basically be hobbling your own ability to give an accurate account of emotions (etc.). And no doubt, far worse, end up thinking of yourself as a far narrower, less complex, less interesting, dumber thing than you really are.
Indeed, I wonder if we might consider there being something kinda intellectually offensive in this supposition. Here's my silly trinket, now, everything is just like that! End all science, we're done boys -- it's just P(Y|X)
Why should I discount a theory just to protect my ego? We've read countless stories about science only progressing when those with big egos die. It would only seem logical that eventually it will come for my own.
mjburgess|2 years ago
to be angry is for your sensory-motor system to be primed for aggression; it's for your cognitive systems to be narrowed and focused on analysing high-threat parts of your environment; it is for your memory-formulation to be modulated towards threat recollection etc.
Sure, if an LLM's prompt "be angry" causes it to adopt a threat stance to its environment, to regulate it's theory-of-mind to engage with possible hostile entities, and so on --- then yes, when LLMs are there, I shall concede the point
However, how terrible it would be to start with an analysis of emotions in terms of the capacities of LLMs -- right?
Since if you did that you'd basically be hobbling your own ability to give an accurate account of emotions (etc.). And no doubt, far worse, end up thinking of yourself as a far narrower, less complex, less interesting, dumber thing than you really are.
Indeed, I wonder if we might consider there being something kinda intellectually offensive in this supposition. Here's my silly trinket, now, everything is just like that! End all science, we're done boys -- it's just P(Y|X)
kelseyfrog|2 years ago