[nosseo]

Hey HN, I'm the author of this article (also the precursor predicting this, which was on the front page yesterday). My impression is that the best place to look for an explanation is actually the facebook post by Luciano Floridi: https://www.facebook.com/floridi/posts/10157226054696031. My sources at Google just couldn't see the panelists constructively working together on a panel at this point. Obviously, protests by Google employees played a role too.

[geofft]

> My sources at Google just couldn't see the panelists constructively working together on a panel at this point.

Well, that's the point, isn't it? If the AI ethics council is meeting four times a year for a couple hours at most, and and has people who can't even agree on "Which bathroom should this person use," how are they going to produce productive advice for actual hard questions that haven't been well-explored?

There is a place for debate between people who don't agree on worldview. This council was never going to be it.

[qmanjamz]

It seems to me that if they limited their recommendations to only things the entire board could agree on, they could still be productive.

For example, they may not agree on how bathrooms should be organized, but they probably all agree that bathrooms should exist.

The fact that the board had wildly varying opinions is a feature, not a bug. The outputs of this board should be widely agreed upon. If they make declarations that half the country disagrees with, they're going to lose credibility fast. Having an ideologically diverse board helps to secure their credibility.

[educationdata]

Is "Which bathroom should this person use" an easy question? If it is really easy, why it is so controversial? You may think it is an easy question, in reality it is an actual hard question.

[gubbrora]

Why should they agree? They could have produced advice of the form "from perspective A we should do x but from perspective B we should do y".

[mc32]

Presumably there would be many more issues beside bathroom policies. Moreover one person isn’t s majority among eight.

[yalue]

You're right about all of that, but shouldn't a meaningful board at least make some show of trying to include diverse opinions? The key word of course is "meaningful", and I think you're right on the money that Google wasn't interested in having a panel to discuss meaningful issues in the first place; it seems more likely that they wanted to be able to point to a source of "external" "unbiased" opinions to justify what they were planning to do already.

(Obviously the people who complained about a prominent right-wing figure would just argue something hypocritical like "only the right kind of diversity of opinion", but that's beside the point; the panel was a dumb idea to begin with, coming from a group with so much of a stake in trying to advance AI for the sake of advertising)