interestiing. i had thought workshop labs' work (training an ai model to amplify each user) was ~lower-priority to me than to them because "how different are we from each other really", but maybe it's also because "we humans have biases we really don't want amplified"
Oh interesting, i've not heard of Workshop labs before. i'm also not sure if i agree with the premise mainly for the second reason you've listed. But does seem like there are some vectors linking intelligence to WEIRDness (at least in AI models) - Musk did try to train xAI to be non-woke only for it to be increasingly liberal?
It does seem to me that 'egalitarianism' is a better "upgrade" to the human cultural software, and not sure if I would want that removed depending on who the base model is interacting with. That seems more of Claude/Anthropic direction than Workshop labs direction?
interestiing. i had thought workshop labs' work (training an ai model to amplify each user) was ~lower-priority to me than to them because "how different are we from each other really", but maybe it's also because "we humans have biases we really don't want amplified"
Oh interesting, i've not heard of Workshop labs before. i'm also not sure if i agree with the premise mainly for the second reason you've listed. But does seem like there are some vectors linking intelligence to WEIRDness (at least in AI models) - Musk did try to train xAI to be non-woke only for it to be increasingly liberal?
It does seem to me that 'egalitarianism' is a better "upgrade" to the human cultural software, and not sure if I would want that removed depending on who the base model is interacting with. That seems more of Claude/Anthropic direction than Workshop labs direction?