The case for value alignment through multiple AI models
A commenter argues that the result reflects the combined values and mental models of the people whose language models receive a prompt.
TLDR
A commenter calls sending a prompt to multiple language models a “great strategy for value alignment.” Their reasoning: the returned result represents the “sum total” of the values and mental models of the people whose models received the prompt.
Combined views
755
1 Source, first seen 15d ago
likes