This was a nice read, I like how you point out that human morals are still pretty bad. That raises a lot of questions on how to align AI’s, and foreshadows that some models will likely be given morals that not everyone agrees with or understands. Lots of challenges ahead.
Thanks Gabe! Yes, I agree that aligning to the right values in the right way which will be widely accepted as legitimate is a pretty deep and broad problem.
This was a nice read, I like how you point out that human morals are still pretty bad. That raises a lot of questions on how to align AI’s, and foreshadows that some models will likely be given morals that not everyone agrees with or understands. Lots of challenges ahead.
Thanks Gabe! Yes, I agree that aligning to the right values in the right way which will be widely accepted as legitimate is a pretty deep and broad problem.