Also, If AI researchers truly grasp point no. 1 - they will not use “language” or “large bodies of text” as the main source of data for reinforcement learning (like chatGPT) they would rather—focus on capturing ethical behaviour or large bodies of information that can capture ethical norms. Not all data (books, internet info etc.) can infer our best characteristics as humans.
Also, If AI researchers truly grasp point no. 1 - they will not use “language” or “large bodies of text” as the main source of data for reinforcement learning (like chatGPT) they would rather—focus on capturing ethical behaviour or large bodies of information that can capture ethical norms. Not all data (books, internet info etc.) can infer our best characteristics as humans.