It’s a really common refrain here that if you are employee critical of your employer you should expect to get fired…
Well perhaps, but you’d think that specifically someone in AI ethics needs to be empowered to speak up against both internal external issues. Or did google just want an research ethics hit team against other companies technologies?
Makes me wonder if project zero is really trying as hard as possible with googles own products. Clearly it’s impossible to be an independent team at google.
It's not just the employer. James' Damore's memo silenced a lot of the more conservative thinkers, especially as he didn't even want that memo to be public, just a research on a mailing list. Harassment from the political left was very strong at Google, that it was just a matter of time until it bit back.
As for me as a white man although I was just doing my job silently, I didn't feel appreciated and tried to isolate myself from meetings / politics and just focus on work (and was happy to leave 1 year after that memo). The fun old times when we could just focus on making the user happy and talk about doing cool stuff was over.
Maybe you are right, I was working in Switzerland, not in US.
What I remember clearly though is that Google's training contained material about how I am allowed to socialize with coworkers outside work, which may be allowed in the US, but it was clearly against the Swiss constitution (my job security was more important to me than speaking up about this though)
What exactly did this training say? I'm just a little suspicious that Google was saying you couldn't socialize with friends outside the office, since that's unenforceable, not in their interests, and also deeply stupid.
I didn't write that that we couldn't, it was about how we could.
For example we shouldn't tell a coworker that he/she's pretty, shouldn't get drunk together, and a lot of behaviour that is considered flirting and getting to know eachother in Switzerland.
In US behaviour outside Google is handled by the company as well, but in Switzerland there are very strict privacy laws and culture that forbid companies to control any part of the life of people outside work. The right to privacy is part of the Swiss constitution, and it's taken there very seriously (to the point where generally the tax authorities don't have the right to look at your bank accounts, unlike in US for example).
Edit: I can't reply, so I extend this comment
,,shouldn't'' is not banning.
And also what you write that Google may have a problem if harassment happens is true, but I still think you think in US law terms: privacy in Switzerland is like voting right in US: it's like telling that ,,you are not advised to vote for Trump as a president''. Or telling you that ,,you are hired, but you should probably go back to Mexico''.
warkdarrior: Again, yes, in US that training is legal and socially acceptable. But the training was in violation of the Swiss constitution...I'm not sure why people in US don't understand that the laws are different in different countries (people from other countries understand it generally). Harassment is a trendy topic right now, but it doesn't trump privacy and the right for not controlling any part of a person's life outside work, even if that's with coworkers, and even if it may lead to a legal problem for Google. In Switzerland people are grownups, they don't need babysitting.
I have to add one more thing: Andy Ruby belongs to prison, everybody knows it, and giving us more trainings won't help the main issue, that the top management was/is raping women or coercing them to have sex to keep their jobs.
Typically that type of employee training in US is meant to limit sexual harassment, where a person thinks they are flirting and the co-worker on the receiving end is offended. HR would step in at that point and take action against the harasser, so i see that kind of training about being careful in socializing as a way to warn people about the consequences of their actions.
Is it sexual harassment because it's in the workplace or it's because the receiving end didn't like it/the flirting person or because flirting IS a sexual harassment?
TBH I am again skeptical that there was a training that said this. I could buy if there were a training that explained that such behaviors could be construed as harassment under certain circumstances (e.g. if you had previously made advances and been turned down). I could buy if there were a training that told you you should think twice before doing these sorts of things. That's just sensible advice, since if someone does take your behavior as harassment and took the complaint to HR, then you're in a bad situation, at best. A training that banned you from doing those things? I doubt.
Corps have a habit of buying in US services tailored for a US audience and then just blithely pushing that crap out to all their international employees with scant attention about outcomes, its just "mandatory".
As such you have to watch some real thoughtless, poorly put together, bottom of the barrel crap so legal can tick a box that diminishes their responsibility.
Given the motivations at play here I wouldn't be surprised if they ballsed it up like this.
I work at Google in California. Unless you have completely different trainings, I don't think your characterization is accurate. For one it doesn't say at all that you should not get drunk together. In fact there are many company events where alcohol is served and some teams keep a stash of liquor in the office.
I work a corp that dishes out mandatory training and I feel like you have the wrong end of the stick. Its not necessarily well put together by smart people, rather its cheaply produced, so this outcome would not be surprising.
So far every example I ever saw of a conservative complaining about their freedoms and speech being limited, they either don't disclose the full story, or when they do, I never end up agreeing that any valid right of theirs was violated or infringed. Instead it's always just crying about having to be even the most basic level of civilized.
And I'm very much in the Bill Maher camp when it comes to liberal counter-productive hyper righteousness and intolerance. And I'm a leftie who disagrees with the left on guns. Try being that guy at parties.
Yet somehow I don't feel beaten into submission and into hiding my opinions for survival, even the unpopular ones. There must be some mysterious other ingredient involved with these poor hobbled heros. What could it possibly be?
Haha, what are you talking about? Are you saying Swiss employees are legally not allowed to socialise with work colleagues? Do you honestly think this happens in practice?
I was confused by GP until I reread it and realised that the operative word is how.
I don't know anything about Swiss law or Google internal policy but I think what he's saying is that it's illegal in Switzerland to dictate how your employees socialise with each other outside of work (and perhaps to dictate anything else about how employees behave in their private lives?), not that it's illegal for employees to socialise at all.
> It’s a really common refrain here that if you are employee critical of your employer you should expect to get fired
I don't know how common that refrain is but I have one a little less controversial: if you use corporate email to exfiltrate confidential documents or coworkers' personal information, you really should expect to get fired.
I don't think that's exactly what happened. From my understanding she was backing up emails that served as evidence to support the claim that Google had treated Timnit Gebru unfairly during their termination. Not sure where that legally falls but I wouldn't exactly paint it in the light you did here.
The legal way to handle that is for Dr. Gebru to sue Google and get the emails through discovery. (If Google didn't do anything that Gebru can/is willing to sue over, well, she doesn't have any legal right to the emails.) Exfiltrating confidential documents will get you fired, and saying you did it to assist a friend in badmouthing your employer on Twitter is not going to help matters.
> legal way to handle that is for Dr. Gebru to sue Google and get the emails through discovery
The not-quite-legal but defensible way might also involve openly forwarding those emails to her lawyer. Sending them anywhere else removes the shadow of doubt one needs when strategically but willfully violating a contract.
Multiple external accounts could mean her lawyer's account and her own. Or 2 accounts at her lawyer's firm. So we don't know if she sent them anywhere else.
IANAL but I wonder, would this be a reckless thing to do without encrypting the data first?
In any case, I would not do this in any form unless my lawyer said it was a good idea first.
But I would also not post a poorly worded rant, with strange footnotes, trashing my employer, on my employer's own document-hosting service, and publicize it on Twitter as soon as I suspect I'm in trouble, so what do I know.
[Edit:] I mean reckless because the email could theoretically be intercepted in transit.
The assumption that corporations, not courts, decide legality seems inane to me. The company you worship could commit a crime and then engage in a cover up to destroy all evidence before discovery could occur. The easiest, recent example is Ebay:
If there's a law that permits exfiltrating confidential documents from an employer because you don't like one of their policies, I'd be happy to know of it. But I doubt such a law exists.
If they were confidential documents and she signed an agreement indicating she would not expose or retain such property, legally, google has standing to fire her and sue her. But morally? She and others may feel it was the right choice, particularly if she felt that Google would not provide the evidence through the discovery process.
It’s not like she’s an ombudsman for Google. If she has concerns she should voice those concerns internally. If she doesn’t feel that her concerns are being taken seriously, she should quit. On the way out the door, she should tell her story.
It’s honestly just basic professionalism. Don’t publicly trash your employer.
I'm curious how you think someone who is a professional ethicist is supposed to do their job without being allowed to comment on ethics.
It's not a cog-in-a-machine profession. Having critical opinions is what the job is about.
It's literally stupid for a corporation to employ an ethicist and then somehow be outraged and appalled when they have their ethics questioned and challenged.
Wanted to write this exact same reply before I saw yours.
She was not a general ethicist at google. Her job was not to comment on the ethical issues of employment. Her ethics role was AI specific and none of this has anything to do with AI
I'm curious why you think she was fired for having critical opinions as opposed to grossly unprofessional conduct. She could have done her job just fine without throwing her colleagues under a bus, badmouthing her employer in the press, and having a paranoid meltdown because her paper didn't pass peer review. Being an ethicist doesn't give you the right to act like an asshole.
>> It’s a really common refrain here that if you are employee critical of your employer you should expect to get fired
I was just critical of my employer, I called them out on how evil they were, I cost them money, I embarrassed them publicly, and then, all of the sudden out of nowhere I was fired!
That's sort of your assertion, right? Am I misrepresenting anything?
Yeah, why wouldn't a company want a workforce full of yes-men and engineers who care about nothing but making shareholders money and making executives look good?
There's a gulf between "hire only yes-men" and "hire activists who openly state a desire to undermine the company, who break its policies and share its internal documents to forward that agenda, and generally spend their global-0.1%-level salary accusing the company of actively harming the least among us".
that's clearly a strawman interpretation of my comment.
> why would a company pay someone to publicly criticize them?
by all means, a healthy company should encourage internal critiques. but if you're going to take that criticism public, you can't expect to also collect a paycheck at the same time. do you make a habit of paying people to work against your interests?
I'm seeking to normalize community accountability practices that are pretty core to many indigenous cultures, including my Irish ancestors.
So, yeah...I'll pay for that in every organization I get involved with from here on out.
Just cause it's not normalized doesn't mean it isn't useful. The reason to do it is for the sake of actual accountability, as strictly internal accountability is nonsense.
This is the business version of "keep it in the family": it gives all kinds of toxic behaviors a shade to hide and grow in, while keeping the public from actually choosing who gets their business based on their values.
"community accountability practices" just sounds like mob justice. It seems like an especially bad idea in the current environment where people who are more than happy to join in an outrage mob after only hearing one side of the story, and sometimes aren't even affected by whatever they're targeting. A "community" shouldn't be anyone who happens to agree with you on twitter.
I think what you're speaking to is the result of a lack of normalized practices actually oriented toward accountability and healing, rather than shaming, canceling, mobbing, etc.
I'm not even saying I know how to do it. I think the current environment is an indicator of a greater need for true accountability and not the theater of "justice."
I'd also say outrage mobs are the direct result of people confusing themselves with who is angering themselves. Many adults I know still insist it is other people who make them angry, instead of it being themselves who use the stimulus of others' actions to anger themselves. We need better EQ/CQ education and what you're highlighting points directly to it.
At this point a fair proportion of the planetary population is affected by FAANG's ethical stance on ad tech. It's looking likely AI is only going to make that worse.
A "community" should also not be an enforced happy smiley corporate PR face.
And mob justice can also look a lot like running people out of town because they're not properly respectful.
But how can community accountability practices work in an organization like Google, which hires from many different communities? It seems like it inevitably produces loud, angry, unresolvable conflict whenever your community and my community don't agree on a decision.
I think the difference lies between harm and disagreement. How can it work? I don't know or have all the answers, as I've never even seen community accountability in action at a small local scale.
This highlights to me the importance of learning how to do this stuff.
I don't think it's as complex as you're making it. Community accountability works through shared cultural standards and social pressure to conform to the status quo; Google employees don't have shared cultural standards, and the whole problem here is that some of them don't think it's right to conform to the status quo. The reason you've never seen it in action is that, in a multicultural society, it doesn't work on any scale larger than a friend group.
It is literally the job of an AI ethicist to highlight ethical issues, so if they're publicly criticizing you, then you're likely doing something unethical.
If your team of ethicists has repeatedly highlighted issues and you continue to not address them, what should they do then? In the case of a company whose self-image is of openness and directness and benevolence, it seems reasonable that publicly highlight the lack of change is within reason.
If they cared less, they'd just shut up, but employees seem to have a higher opinion of Google than Google does of them.
Firing them in the middle of that suggests an unwillingness to deal with the previous issues, doubling down on unethical behavior.
I might be speaking from ignorance here, but I highly doubt their job description said anything about making accusations via tweet.
> If your team of ethicists has repeatedly highlighted issues and you continue to not address them, what should they do then? In the case of a company whose self-image is of openness and directness and benevolence, it seems reasonable that publicly highlight the lack of change is within reason.
if they strongly disagree with the direction the company is taking, they ought to refuse to be part of it and resign. public criticism can wait until they stop collecting the (quite fat, I imagine) checks.
edit: I guess I'm not really objecting to the public criticism before leaving the company. I'm objecting to the incredible feeling of entitlement it must take to expect to keep getting paid while actively damaging your employer.
And how exactly is your suggestion going to play out in the long term?
All the critical employees resign, until they are all eventually replaced yes-men who just care about the money or are too afraid to speak out.
And if a company is doing something unethical, then why is it inherently _bad_ to damage that company until it changes direction? Acting in unethical ways has to hurt the bottomline, otherwise what incentive is there for a company to change?
I don't understand why people seem so confused about the notion of employment.
If a company hires a security researcher, and that security researcher finds that the company is not listening to their advice, is it their responsibility to publicly expose the security flaws of the company?
What if you hire a plumber, and the plumber doesn't find that you're taking his advice regarding your toilet. Is it the plumbers responsibility to publicly comment on that?
Well, what she is doing is called performative acitvism. She is doing staged martyrdom in front of Twitter mob and activist media. She can try to do the job well, if she felt like she was being hired as a PR move and not for any actual impact, she should've quit and told her story. I guess she didn't feel she will get any media attention if she has taken the high road. Instead she created all these drama including blatantly violating company policy. I am not sure whether she has done all these things in good faith.
Unfortuantely the amount media attention (including this very thread) these people get, it will encourage more staged martyrdoms. Soon, we will have our own tech industry reality TV show. Sadly, that's the reality we are heading into right now.
I agree that employees should be allowed to criticize their employers, but there's nothing inherent to an Ethics specialist that would weight things even more in that direction. Ostensibly, she was hired to report internally on ethics considerations. Unless part of her job description was to liaise with the public on those ethical issues, public exposure would come as a bit of a surprise to her employer.
Well paying someone to bite yourself is never a good sale in private organizations.
Otherwise, why regulation, like at all?
Google is a giant corporate with deeply vested interests all around. It is laughable that it still tries to put up an ethnical makeup to just feel good about itself. Which inlines with my observations that Googlers often have this narcissism rooted deeply in themselves, the eagerness to appear to be better.
Most activism isn't successful. You would find your way to violence very quickly if a few examples of unsuccessful non-violent activism caused you to justify breaking more and more serious laws each time.
Google decides for your behavior. And you decide for Google.
The person giving you money is allowed to decide whether they want to continue giving you money. Similarly, you are allowed to decide whether you want to continue taking the money.
If google is acting unethically then I, as a user of google, want to know so I can hold them accountable.
Otherwise having an “ethics committee” is just feel-good marketing signaling at the same level of privacy terms stating how google, the ad behemoth, “cares about our privacy”.
> but you’d think that specifically someone in AI ethics needs to be empowered to speak up against both internal external issues.
I'd think ethical (AI or not) researcher would work far away from Google's influence as possible if they really want to analyze Google AI work.
> Makes me wonder if project zero is really trying as hard as possible
I wouldn't be surprised if project zero find as many vulnerabilities as possible in Google's software and give them a chance to fix quietly.
If there are huge security problems to be found in their software then independent researchers can make big name/fame and money by publishing them out.
It's not that I want them to get fired, I'm just aware that the whole point of an in-house ethics committee is to excuse the objectionable stuff, not to publicly criticize it. Sort of like when people complain that HR wasn't on their side, but actually just ratted them out to management. Yeah that sucks, but that's their role. Be smart about how you dissent. Be anonymous and leak stuff. You'll still get to keep on cashing that paycheck and probably have a more meaningful impact on the broader world compared to some de-fanged "ethics" committee press releases.
It seems like Google really likes firing people instead of free speech. Sure they can do that because of at will employment but at some point it has to start leaving a bad taste in everyone's mouths especially when these firings get publicized. At some point CS grads will start taking this into consideration when it comes to working at places and having their own political views. Who wants to work at a place where there's a witch hunt? This isn't just leftists getting fired either
Maybe when Google stops paying these CS grads obscene amounts of money. Only people with very strong political views will convince themselves to not go there, and I doubt it's a large group of people.
> It seems like Google really likes firing people instead of free speech.
Because it has consequences in the outside world that need to be smoothed over. Some people think that Google should just let people do whatever and allow themselves to be the whipping boy for what their employees have said.
Makes me wonder if project zero is really trying as hard as possible with googles own products. Clearly it’s impossible to be an independent team at google.