News for hackers. Stuff that matters.
Re:Google’s new AI ethics board is already falling apart by hirundo on Wednesday April 03, 2019 @09:00PM (#19568514)
Really, or is it just about where you'd place the Overton window? Should they include a white supremacist, pedophile, human extinction advocate, etc? If so you're being consistent. If not where would you draw the line? By the popularity of the ideology?
Re:Google’s new AI ethics board is already falling apart by pdonis on Wednesday April 03, 2019 @10:34PM (#19568899)
> Should they include a white supremacist, pedophile, human extinction advocate, etc?

In an ethics discussion? Short answer? Yes. Otherwise how can you be sure that their beliefs don't have some merit? Remember that "sure" here means "sure enough to bet the entire future of the human race on it", because that's what's at stake with AI ethics (unless it turns out that AI is impossible, which I don't think is likely and in any case I wouldn't want to bet on it). No human being knows enough about ethics to be able to rule out any viewpoint with that level of confidence.

Re:Google’s new AI ethics board is already falling apart by naasking on Wednesday April 03, 2019 @09:57PM (#19568741)
There might be a few criteria. Some balance of representative of the population, internal and factual consistency, perhaps even length of the tradition. That is, a candidate wouldn't have to have all of these, but satisfy at least some of them to some extent.

I think would exclude the most objectionable, but you have to allow for the fact that it will include prominent beliefs that you won't like. Sorry, most of humanity literally doesn't share your point of view on many topics.

Nor should such a council endeavour to create an echo chamber.

Re:Google’s new AI ethics board is already falling apart by mike00632 on Thursday April 04, 2019 @05:56PM (#19577503)
>Nor should such a council endeavour to create an echo chamber.

But if you create a council and staff it with viewpoints that only serve to ignorantly speak in favor of tradition and cultural norms then haven't you just created an echo chamber?

Re:Google’s new AI ethics board is already falling apart by naasking on Friday April 05, 2019 @11:39AM (#19583604)
Sure, if you accept your qualifier of "only". I see no reason to though.
Re:Google’s new AI ethics board is already falling apart by solotronics on Wednesday April 03, 2019 @09:45PM (#19568688)
If you don't include people you disagree with politically they will go off and form their own group independent of yours. I can imagine the defense contractors banding together and making a new Manhattan Project for general AI. Wouldn't it be better to have an ethics board that includes all the big players?
Re:Google’s new AI ethics board is already falling apart by gowld on Wednesday April 03, 2019 @09:48PM (#19568700)
Either they'd do that anyway when they didn't get whichever killbot they wanted, or else they'd convince you to build the killbot you wanted, and in either case you end up with the killbots you don't want, so inviting them to your ethics board doesn't help.

The defense contractors haven't invited you to their ethics board.

Re:Google’s new AI ethics board is already falling apart by sgt101 on Thursday April 04, 2019 @03:10AM (#19569951)
Darpa are keenly looking at this. They have to because its important that such advances would change the strategic situation. But, let's be honest, it's science fiction. There is no prospect of aig at all in our life times.
Re:Google’s new AI ethics board is already falling apart by bigj0n on Wednesday April 03, 2019 @10:50PM (#19568938)
I think that you should hear them out and they would be very easy to dismiss very quickly.

The reason we dont agree with White supremacists isnt that we've effectively suppressed their speech. It's that they're wrong and it's simple to see.

Re:Google’s new AI ethics board is already falling apart by woodruffw on Wednesday April 03, 2019 @10:59PM (#19568976)
This is something that continues to confuse me: what distinction do you draw between "not hearing out" and "dismissing very quickly"? It's not like none of us know what a white supremacist is -- we've all seen their rhetoric, have dismissed it, and no longer feel inclined to re-litigate the issue.

Put another way: you wouldn't, presumably, ask a physicist to critically re-evaluate the laws of thermodynamics each time they perform a calculation. Why are you holding ethicists to a higher (and progress-impeding) standard?

Re:Google’s new AI ethics board is already falling apart by elliekelly on Wednesday April 03, 2019 @11:55PM (#19569188)
This is going to sound silly but I think there is pretty substantial value in making the white supremacist feel _heard_ and then feel dismissed. How do you feel when someone brushes off your opinion as invalid before you've even had a chance to express it? Like the person brushing you off is obviously an idiot who can't be reasoned with. But if someone has taken the time to discuss something with you and really hear you out and then they dismiss your opinion? Then you're probably a little more open to thinking through the merits of why they disagree.

I'm not sure how to strike that balance and I completely understand that it feels both distasteful and inefficient to "re-litigate" things like white supremacy when the correct outcome is obvious to us but I think the negative effects of "echo chambers" are exacerbated by how quickly their views are dismissed before they've been given a chance to explain themselves. In denying them the opportunity to express their views we're not only solidifying their views but denying them the opportunity to examine their views critically.

Re:Google’s new AI ethics board is already falling apart by woodruffw on Thursday April 04, 2019 @12:22AM (#19569282)
That doesn't sound silly at all to me. You bring up a very good point.

However, I also think you give the white supremacist too much credit. He wants to be heard, but he isn't a faithful member of the conversation. He can't feel dismissed to, because (from his perspective) it's nobody's place to dismiss to him.

His goals are simple: (1) to keep people talking about white supremacy, (2) to frame his removal from the conversation as a violation of the principles that he himself seeks to undermine, and (3) to recruit more people into white supremacy via (1) and (2). He's a door-to-door salesman for hate: he's here to sell you something, not to critically examine his views. He knows that conversations with him won't work, and that's exactly why he wants us to think that he's worth conversing with.

Edit: I'm not going to modify what I wrote above, but I want to condition it a bit. I should really have said that he isn't always a faithful member of the conversation. When he is, all of your points apply. My overwhelming concern is that we give him the benefit of the doubt too frequently.

Re:Google’s new AI ethics board is already falling apart by mantas on Thursday April 04, 2019 @12:54AM (#19569437)
Now that's a nice strawman.
Re:Google’s new AI ethics board is already falling apart by naasking on Thursday April 04, 2019 @08:57AM (#19571670)
> It's not like none of us know what a white supremacist is -- we've all seen their rhetoric, have dismissed it, and no longer feel inclined to re-litigate the issue.

I actually don't think a lot of people know what a white supremacist is after seeing who has been slapped with that label in the past few years. The rhetoric around a lot of these issues is very disingenuous these days.

So while you're correct that white supremacy is an argument or philosophy that is easily dismissed, but who gets that label is very contentious indeed even though it seems like it should be straightforward.

Re:Google’s new AI ethics board is already falling apart by WalterBright on Thursday April 04, 2019 @02:11AM (#19569725)
> ask a physicist to critically re-evaluate the laws of thermodynamics each time they perform a calculation

Perhaps not. But it is a good idea for scientists to periodically re-visit their assumptions about how things work. Like what Einstein did.

Re:Google’s new AI ethics board is already falling apart by ralusek on Thursday April 04, 2019 @03:25AM (#19570019)
Presumably part of the case on their behalf would be that most or many of the people accused of being white supremacists, or afforded other such titles, are simply not. Granting Ben Shapiro, Jordan Peterson, et al, an opportunity to speak is giving people accused of being white supremacists the opportunity to not only quickly prove that their positions are not collectivist, let alone racial, but (more importantly) represent valid counters.
Re:Google’s new AI ethics board is already falling apart by tremon on Thursday April 04, 2019 @05:14AM (#19570427)
Here's what I think is the problem with that approach: these people will use "you hearing them out" as proof (to their base) that they are right. These people will ignore your dismissals. These people will use every possible trick, even the ones not in the book, to misconstrue your actions to establish themselves to their own base.

They should not be given a platform. At all. And even that will be misconstrued as "see, we're being marginalized because we're inconvenient but true".

Re:Google’s new AI ethics board is already falling apart by MockObject on Thursday April 04, 2019 @03:09PM (#19575982)
> these people will use "you hearing them out" as proof (to their base) that they are right.

Their base already believes they're right. If anything, the deplatforming will be used as further proof (to their base) that they really are the victims of, in this case, a powerful Jewish conspiracy.

Re:Google’s new AI ethics board is already falling apart by naasking on Thursday April 04, 2019 @03:30PM (#19576211)
> They should not be given a platform. At all.

They should have access to the same platforms that are accessible to the rest of the public, so long as their speech is not an incitement to violence.

Protest those ideas if you like, but living in a democratic society means battling with words and not censorship.

Re:Google’s new AI ethics board is already falling apart by mike00632 on Thursday April 04, 2019 @06:01PM (#19577547)
Having a seat on Google's ethics board is not a platform accessible to the rest of the public.
Re:Google’s new AI ethics board is already falling apart by naasking on Friday April 05, 2019 @11:40AM (#19583617)
Sure, but that's not what the other poster said. They said they shouldn't have a platform "at all". Meaning, "any platform", which is the point I disputed.
Re:Google’s new AI ethics board is already falling apart by tremon on Tuesday April 09, 2019 @05:02AM (#19612919)
To be entirely clear, I said "be given". I said nothing about them arranging a platform for themselves.
Re:Google’s new AI ethics board is already falling apart by naasking on Thursday April 11, 2019 @05:32PM (#19639454)
Firstly, "arranging a platform for themselves" is the same as "being given a platform in exchange for goods and services (typically money)". If you're making this argument, then there's no distinction in what you describe. They're "given a platform" in exchange for services rendered as an AI ethics panel member.

Secondly, it's debatable whether they're even being "given a platform" in this case anyway, because their duties are specifically surrounding AI. Going off on an anti-trans rant or something would get them removed. Since that's the sort of platform you're presumably arguing against giving them, so there seems to be no conflict here.

Presumably, what you actually meant by "not giving them a platform" is that they shouldn't be given legitimacy or validation. I'm not sure I can agree that their presence accomplishes this. Some people give the pope some moral authority, but plenty enough people aren't swayed by his fancy hat, and his lofty moral status enables good discussions over what's truly ethical. Christian and Catholic engagement is and has been falling precipitously.