[sdiy] AI code generation

[mskala at northcoastsynthesis.com]

Among many problems with getting involved in a discussion of environmental
impact of AI, is that for very many people who participate in such
discussions, environmental impact is not the real issue.  It is at best a
rationalization of a position taken for other reasons.

Recent research at the University of Toronto supports the idea that many
opponents of the use of AI, oppose AI as a moral issue - seeing AI as
wrong *in itself*, not because of any consequences or claimed
consequences.  The distinguishing feature of that kind of opposition is "I
say it's bad for the environment, but even if that were proven not to be
true, I would still oppose it."  So why bother even discussing whether the
consequence-based point is true?  Is it to convince others, with an
argument that didn't convince oneself?

I did a video going through the research paper in detail on my site here:
   https://matthewexplains.com/11190144/

Most videos on that site require at least creating a free account to watch
(and many are pay-only), but that particular one is free without login.

For either side I think it's valuable to know as much as possible about
what's really being discussed, and this paper seems like a step in that
direction.

-- 
Matthew Skala
North Coast Synthesis Ltd.