I'm fired: Google AI in meltdown as ethics unit co-lead forced out just weeks after coworker ousted Google has finished its probe into the controversial ousting of Timnit Gebru, co-leader of its Ethical AI unit. The ad giant promised to implement new procedures around “potentially sensitive employee exits,” though it did not make its findings public. Gebru said she was fired for warning coworkers in an internal memo that, …
Point 1: Google fucked up here, big time
Point 2: Don't ever email your employer with "...here are the conditions. If you can meet them great I’ll take my name off this paper, if not then I can work on a last date." - that's offering to resign. Make them fire you.
it's not a resignation letter by any stretch of the imagination.
No, but it is a resignation offer: "change this, or I'll leave". Not really surprising if the response is "No. Sorry to see you go". Leaves the sender with little choice but to leave, or eat humble pie & agree to stay anyway, with very well-burned boats behind them.
There was a case in the UK many years ago that I remembere reading about where a Nurse thought that a collegaue wanted to get rid of her so was making her job as difficult as possible. The Nurse responded by writing a letter to HR to make a complaint about the way she was being treated - however, she added an extra comma into her statement that "xxx is making my job a misery, so I resign". The HR dept interpreted this as a resignation letter and ended her job. She responded saying that she hadn't resigned and her letter had been misinterpreted and the case went all the way to an industrial tribunal which ruled in favour of the HR dept on the basis that the letter was (gramatically) a resignation and that was how it was legally to be interpreted whatever the writers intentions had been.
> Point 2: Don't ever email your employer with "...here are the conditions.
1. Not really. Not in the USA.
2. You don't know the details of the email, as it has not been made public.
3. Unless you work for Google HR, in which case I know why you're here.
Actually, retaliating against an employee who asked their employer for better workplace conditions is grounds for filing a formal complaint with the NLRB.
Google is the company that was trying to prevent its employees from discussing compensation with their colleagues. That is blatantly illegal as well.
It looks like Google wants to catch-up and soon surpass Amazon in terms of workplace toxicity.
Q. What's the difference between responsible and ethical?
A. Behave abominably, and you're responsible, but you ain't ethical.
This has turned into a complete farce, due to anybody honestly concerned about ethics coming up against the corporate wrath of an unethical company.
It's hard enough doing "ethical" research at a university - you'd hope that someone working in that field would not be so naive as to be blind to the potential hazard of being on the payroll of a company with a vested interest.
Nothing from Google is truly 'free' and that includes speech.
Copying company materials onto unsecured and unauthorised networks and devices is a breach of professional ethics. Including HR information (as alleged) makes instantly sackable even when the intent is good and in the company's interests.
"Copying company materials onto unsecured and unauthorised networks and devices is a breach of professional ethics. Including HR information (as alleged) makes instantly sackable even when the intent is good and in the company's interests."
It's certainly a breach of corporate policy. Whether it's a breach of ethics depends on the (ex)employee's motive for doing it. If doing it for revenge or profit, it's unethical. If doing it as a whistleblower, it's ethical. Either way, it's against corporate interests, so don't get caught. In the first case, don't get caught *ever*. In the second case, don't get caught until the docs are securely in the lawyers hands.
From time working on Clinical Ethics boards, some of the stuff that the pair had been doing would definitely not have passed muster as "Ethical". Ideological crusades rarely are.
There was a whole fanatical wave of "It's all against minorities, it's only because she's a woman, it's an attack against women and minorities" in this last event, well past the time that Google stated that the exfiltration of data contained sensitive information on other people (this is ethically a BIG no-no, and would get anyone in an ethically sensitive role fired post haste).
Behaving ethically is a full time task. It's damnably hard to meet standards all the time every time, especially when you have a desire to do something else, but if you can't be trusted to maintain at least a reasonable standard (don't go playing with other people's sensitive data unless it's what you're doing as part of your job, with an acknowledged 'need to know', and protecting it with everything you can), then I'd probably be of the opinion that you should in no way be responsible for writing ethical policy.
Googlers think they work in an environmental where their specialness will be tolerated.
They're then shocked to find our they work for a giant corporate multinational that doesn't like people biting the hand that feeds them, or people making demands, or people search their networks for information.
They're worker drones, and whatever gender, colour or whatever they are Google *doesn't care*. They don't discriminate but they don't tolerate either.
I think some people want a nice free startup/academic environment and they're in totally the wrong place for that.
A company doesn't need specific employees, they need bums on seats with the right skills and these days it's a lot easier to replace someone who's indicated they want to leave than risk them staying, putting up with their BS and potentially having lots of costly "leaks".
She threatened the company, it responded. Before you say she didn't threaten "Do this or I do this" is a threat especially when there is researched etc on the line.
"I understand we could have and should have handled this situation with more sensitivity. And for that, I am sorry."
Translation: "Oops, you were not supposed to hear about this"
"enact new procedures around potentially sensitive employee exits."
Translation: "Now we'll force NDAs on everyone to top it happening again"
The ad giant promised to implement new procedures around “potentially sensitive employee exits,” though it did not make its findings public
Disclosing proprietary intellectual property secrets for free is no way to make and keep friends secure and safe from harm, and especially so whenever working with potentially sensitive information for advancing intelligence .... and all variations and permutations on that AI theme.
It is not at all surprising its finding and promises to implement new procedures are not public general knowledge.
The Gebru situation seems bad, really bad, for Google.
OTOH, the Mitchell situation seems reasonable. Their automated network security system flagged her for shipping (I can't say exfiltrating with a straight face) mass quantities of data off site which is something any company would act on.
In both though, Google's response was so tone deaf that it is unbelievable.
Process is supposed to be "tone deaf". It takes a situation, works out if things fit criteria, then moves on with taking action.
Both Gebru and Mitchell had done enough to trigger action. Huge companies can't afford to spend endless sums on going through someone's life story, making excuses for them, seeing what the public think of it (when they really don't know, and can't ever truly understand the frame of reference) and so on.
Most people either go to court and claim unfair dismissal if they really believe they've been mistreated, or they quietly go looking for the next job.
These two have both claimed it's because of <insert woke trigger phrase here> and put it out into the media. Which is incredibly harmful to future employability. Unless of course that it's in an ideologically driven group and the ideology is of more importance than getting on with the job and doing impartial work.
I suspect they both thought of Google as an ideologically malleable entity that they could configure from their roles such that it bowed to their agenda as a priority over getting actual work done that was needed in their remit. They weren't correct.
"I suspect they both thought of Google as an ideologically malleable entity that they could configure from their roles such that it bowed to their agenda as a priority over getting actual work done that was needed in their remit."
I suspect instead they thought their role would be somewhat related to their job description(s) in the performance evaluation system. I doubt in the hiring interview Google told them "we need you as a front to deflect criticism we're getting that our AI research is unethical, just tell everyone we're great no matter what you actually see". Creating a new Responsible level to oversee the Ethical level clearly shows the Ethicists have gotten out hand. What's wrong with employees these days, can't they read between the lines?
`The Stimulating Collaborative Advances Leveraging Expertise in the Mathematical and Scientific Foundations of Deep Learning`. I scarcely have to close my eyes to imagine this coming out of Borat's mouth. Could they not spare $50 of the $15m for a grade school student to come up with something a little less ham-fisted? Or, maybe a program to generate one?
Google is the Devil, but...
Apart from not understanding whatever their ludicrous jobs were for --- a research paper on the negative environmental and social effects of massive, power-hungry, biased language-processing neural networks like the ones used by Google. --- nor why such companies need ethics departments. The decline of religion, regardless of whether one is a believer, aside, most people will know the correct action usually [ though hopefully not pandering to temporary moral pronunciation, which does not deal with thieving or cruelties as is traditional in moral codes, but with whatever the speaker wants to be ] and it is fairly easy for a company to stick to the straight and narrow *. Therefore the existence of these moral pundits is in itself otiose. However, if they misbehave --- or test the resignation limits --- the company is entirely within its rights to drop them.
.
* None of which implies the functions of the company, which is after all, advertising, mostly, has any moral merit; but their actions and interpersonal relations within that ambit can be moral.
The Register - Independent news and views for the tech community. Part of Situation Publishing
Biting the hand that feeds IT © 1998–2021
