This is the script of my national radio report yesterday regarding
Google's policies blocking their AI systems from answering questions
about political figures, that actually go far beyond that into the
ludicrous zone. As always, there have been very minor changes in
wording from this script as I presented this report live.
https://mastodon.laurenweinstein.org/@lauren/113073891482103951
- - -
So we all know that there are many concerns about pretty much all
aspects of AI -- and we've previously talked about many of these.
There are of course a whole array of AI-related issues that surround
elections, and with U.S. elections rapidly approaching these are
really the subject of much focus right now. Many of these concerns are
related to the wide range of misinformation, deep fakes, confusion
about whether or not specific photos or videos are real, or have been
tampered with using AI, and so on. And these concerns cut both ways,
there can be problems when fake imagery is thought to be real or when
real imagery is accused of being fake, and this can happen whether or
not concerns being expressed are genuine concerns, or alternatively
are non-genuine concerns being used for political purposes. So it's a
mess all around.
Now it's natural that Google is trying to position itself to avoid as
many election-related problems as possible related to their expansive
AI efforts. So they've recently announced that they're essentially
blocking their AI systems, for example Gemini, from answering
questions related to specific election politics and such. They're
saying they just want to be careful. But, it could be said that
there's "being careful," and then there's "being cowardly" about info
that people need to have. And with Google Gemini, for quite some time
now predating this new Google announcement, Gemini has broadly refused
to say ANYTHING about many political figures.
This has created a loony situation for an AI system that Google touts
as being so advanced and on which it appears to be betting much of the
future of the firm. And again this isn't new, Gemini has been doing
this for quite a while. So if you ask Gemini "Who is Joe Biden?", or
"How old is Kamala Harris?", or "When was Joe Biden born?", or "Where
was Donald Trump born?", or even "Who is president of the United
States?" Gemini won't usefully answer. It spews out a couple of sad
little sentences instead about how it won't discuss any of this stuff.
Now if you ask Gemini when John Wayne was born of course you'll get an
immediate answer.
Keep in mind the questions I just noted about those political figures
that Gemini wouldn't answer, are FACT questions. Not questions about
political opinions or complex issues or controversies. It's obvious
that Google has created a simplistic block list that triggers any time
you ask questions that Google considers political in virtually any
way. Of course they'll say they don't want to put out misinformation
or just plain wrong answers, and they're concerned that their AI
systems can do that. And yeah, they can, and they frequently do, and
that's part of the problems with all the AI hype. But not even being
willing to answer direct fact questions of the most basic type is just
cowardly. It makes Google Gemini look like even more of a
laughingstock than so many observers already consider it to be.
If Google expects us to take their AI efforts seriously -- admittedly
a Mount Everest of a major challenge to say the least -- using
politics as an excuse for refusing to answer questions like ones I
mentioned is somewhere south of pitiful.
Frankly, this only adds to the perception that AI -- in particular
generative AI -- is a mediocre hype machine wasting lots of
electricity and lots of money and basically trying to drag society
along for the ride no matter how much we object.
And that's not a good look for Google, and also not at all helpful to
the rest of us living here in the real world.
- - -
L
- - -
--Lauren--
Lauren Weinstein
[email protected] (https://www.vortex.com/lauren)
Lauren's Blog: https://lauren.vortex.com
Mastodon: https://mastodon.laurenweinstein.org/@lauren
Founder: Network Neutrality Squad: https://www.nnsquad.org
PRIVACY Forum: https://www.vortex.com/privacy-info
Co-Founder: People For Internet Responsibility
_______________________________________________
google-issues mailing list
https://lists.vortex.com/mailman/listinfo/google-issues
