Hacker News Reader

They didn't try to do that, though. They tried to organize a group of people who largely (although not exclusively) knew nothing about AI, or ethics, to rubber-stamp their actions and assuage the concerns of regulators & watchdogs. Ethics laundering, basically.

That's a really huge assumption, considering they never met (afaict), let alone deliberated anything or issued recommendations. Google sought input from outside their bubble and those inside the bubble immediately protested based on identity politics.

There are a lot more people outside the bubble than inside, and this action makes it seem like Google doesn't give a shit what they think. It's the sort of thing that makes heavy regulation appealing even for those who normally would oppose regulation on principle.

How does that benefit Google?

There was widespread protest from outside the bubble, too. You'll be unsurprised to learn the entire AI ethics space is being enthusiastically colonized by Thought Leaders all too happy to hold "discussions" in service of whatever their corporate funders desire. Oh sure there'll be some token pearl-clutching, but you can't stop progress (toward ubiquitous facial recognition) right?

Anyway, if you're interested in seeing all the ridiculous shenanigans these people try to pull I recommend following @farbandish on twitter.

I thought it was only 1 or 2 people out of 8 who were problematic; How is that "colonizing"? It also stills feels disingenuous to judge all 8 people's intent and position without them meeting a single time or making any deliberation.

> I thought it was only 1 or 2 people out of 8 who were problematic

I think ahelwer meant the AI ethics space in general, not just this panel.

As a researcher who has been working in ML and also software engineering ethics since long before "AI ethics" became a thing, I've definitely noticed a massive infusion of self-promoting "thought leaders" who have zero expertise in either of those things.

Most of those people are MBAs with little or no actual business experience and zero software knowledge (let alone ML or AI research experience). Their ethics background usually amounts to a few undergraduate philosophy courses that, ironically, failed to teach them the one thing that a philosophy course should teach: a modicum intellectual humility.

And yes, those folks tend to care mostly about their paychecks.

> There was widespread protest from outside the bubble, too

It seems implausible that anyone not in a bubble cares what happens to Google's 8-person no-actual-power ethics team. It could be composed of witches and warlocks who spend their days researching pentagrams and nobody should care. There are a lot of small, questionable teams in large companies.

It is only remotely newsworthy as evidence that Google culturally purges conservative voices. Even with that frame; it could just be a beat up. I'd define the bubble as people who were getting worked up over it; they are obviously culturally synchronized because there isn't an obvious rational link.

Actually I think you'll find quite a few people care about what the megacorps decide to do with AI and their approach to the ethics surrounding it, since it's currently a fantasy to hope for regulation by the US gov.

That is what someone should be protesting, isn't it?

Perhaps the easiest way to help the world avoid the potential problems created by the Google's work is not to work for Google?

These protesting employees, whatever the cause du jour, are all in a compromised position. They need/want their jobs at Google. They cannot effectively function as internal watchdogs and regulators. They are not representative of the wider affected population; we do not elect them.

Neither Google nor Facebook can be relied on to "regulate itself", whether it comes from top-down decision-making or in response to internal protests.

This should be the canary/coal mine for whether or not regular humans can have opinions that differ from the master class/billionaires. Even if you are a titan in the field of AI Ethics, google has no use for you if you disagree with the masters. Google will lose 2 unbelievable female scientists, who also have exhibited far more compassion and empathy then we can expect from googles AI in the future.

Yes they are. It's called "protected concerted activity" and it's protected under the national labor relations act.

These laws were passed because they were seen as better than the status quo, which was employers firebombing their employees or highering mecenaries to shoot them.

Google telling Whittaker that they don't want her doing work on AI ethics after she torpedoed their first effort to have a public discussion is not even close to firebombing her or hiring mercenaries to shoot her. Nor was she engaged in trying to form an employees' union, which I think is the activity protected by the NLRA.

They didn't even fire her. Apparently they just said, We don't want you involved in the AI ethics discussion going forward. That seems reasonable, considering her past efforts in that area have not benefited Google in any way, but rather harmed them.

Just as employers can't firebomb their employees for organizing, employees can't bury steel spikes in logs headed for the sawmill.

Did she "torpedo" their efforts, or merely point out that it was already holed below the waterline, by design?

> or merely point out that it was already holed below the waterline, by design?

That assumes some humility and objectiveness from her higher-ups. "Oh you publicly criticized our project, behind which some of us have staked our reputation. And you created a major negative PR event as well. Thank you for that! We finally see our mistake and we'll apologize and move on". In some hypothetical rational universe it might play out like that, but in reality I don't see how that was supposed to turn out without some repercussions for her.

It perhaps assumes a company which was actively and seriously interested in the area of AI ethics.

Ding ding ding!

The NLRA protects concerted action by employees regarding working conditions. Arguably, this could fall under that definition.

How is an outside ethics panel going to affect their working conditions? The people on the panel don't have any say on employees' pay, promotions, disciplinary actions, assignments, or anything else that might affect their working conditions.

The idea was to have some people from outside the company look at the tech and its potential hazards and provide some input on the ethics of developing and deploying it. People inside the company said, No, we don't want that particular viewpoint to be have a seat at the table on this outside committee. The ethics panel had nothing at all to do with their working conditions.

You are commenting on different things. She pointed out the political oriented people on the ai committee who knew absolutely nothing about technology or ai. They were just prominent conservatives, who had no apparent reason to be there. Pointing out something like that is vastly different than saying you killed your own project. Pointing out what appear to be politically connected but otherwise useless people on leadership groups is a good thing I think.

I don't understand the argument that someone participating on an ethics panel has to be an expert on the tech. The more important thing is that they should be conversant with ethics.

The questions involved have to do with should we build this tech, not how to build this tech. To argue otherwise means that the vast majority of people should not be allowed to have opinions on the ethics of nuclear weapons, medical procedures, or any other topic where they are not subject matter experts.

AI tech and the ethics of using it is rightly a political topic, since it has the potential to affect everyone on the planet.

Preemptively saying that certain viewpoints have no place at the table simply means the political discussion will move from a friendlier environment within Google to the (likely more hostile to Google's interests) environment of general politics in the U.S. Personally I think that might be a better outcome, but from Google's standpoint it's potentially a disaster since it might result in heavy regulation of their business.

You don't have to be an expert, but imho you should be someone who has a grasp of technology, science and the world on a very focused technology issue. If you are someone who denies humans impact the earth's climate, if you argue in favor of people who think the world is only 8,000 years old, I don't trust you to make choices about science based policy.

> the status quo, which was employers firebombing their employees or highering mecenaries to shoot them

Did we not, uh, already have laws against firebombing or hiring hitpersons? I would have thought these kinds of rights-to-collective-action laws are more about employers just summarily sacking people who do things that they're legally entitled to do but that are politically disadvantageous to their employers.

Whittaker organized a walkout to protest the way that allegations of sexual assault were mishandled. Is there evidence that she also organized the petition related to the external AI ethics group in Google?

She was active in organizing the petition drive that successfully pressured Google to keep Kay Coles James off the committee. [0] [1] [2] [3] [4]

[0] https://www.theguardian.com/technology/2019/apr/01/google-ka...

[1] https://www.gizmodo.com.au/2019/04/google-employees-call-on-...

[2] https://dailycaller.com/2019/04/04/leftist-googlers-kay-cole...

[3] https://dailycaller.com/2019/04/05/google-drops-heritage-fou...

[4] https://www.newstarget.com/2019-04-06-leaked-emails-suggest-...

Being asked to abandon a current role is way different than Google deciding to remove said positions. It's the same difference between leaving the company and being fired. Even though in both ways you're not in the company anymore, how it was done is an important factor. Google is basically asking the employee to resign instead of telling her that her position has been removed.

Because she is a prominent female scientist, and she’s proven to be mentally competent to handle this s-storm in an admirable way. It’s a horrible look.