Interesting thanks for clarifying, and yes your rephrasing about ‘in verifiable domains’ is helpful (though obviously we’d want to continue to think about how that verification or validation framework is overseen and by who).
I think I share your concern about the human habit that will slip in of deferring to AIs and this is more of a behavioural psychology issue that we need to address in terms of how humans relate to AI which probably needs interventions on the training, skills and capabilities side of thing across education more widely. Obviously far greater than the domain of animal rights you’re considering here but it’s a useful live example to think some of this through.
I would encourage you and others to not give up too easily on keeping humans in the loop, indeed! But I do think this debate needs a dedicated container to hold it.
Interesting thanks for clarifying, and yes your rephrasing about ‘in verifiable domains’ is helpful (though obviously we’d want to continue to think about how that verification or validation framework is overseen and by who).
I think I share your concern about the human habit that will slip in of deferring to AIs and this is more of a behavioural psychology issue that we need to address in terms of how humans relate to AI which probably needs interventions on the training, skills and capabilities side of thing across education more widely. Obviously far greater than the domain of animal rights you’re considering here but it’s a useful live example to think some of this through.
I would encourage you and others to not give up too easily on keeping humans in the loop, indeed! But I do think this debate needs a dedicated container to hold it.