AI alignment isn’t really OpenAI’s primary mission. They’re seeking to democratize access to AI technology, by developing AI technologies in the open (on Github, etc.) with permissive licenses. AI alignment is sortof a side research area that they are committing a small amount of time and resources to.
AI alignment isn’t really OpenAI’s primary mission. They’re seeking to democratize access to AI technology, by developing AI technologies in the open (on Github, etc.) with permissive licenses. AI alignment is sortof a side research area that they are committing a small amount of time and resources to.
It says right on OpenAI’s about me page:
That as stated looks like AI alignment to me, although I agree with you that in practice they are doing exactly what you said.