Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Tensions in Google's ethical AI group increase as it sends demands to CEO

cnn.com

Post reply on HN
Print view View original

411–420 of 460 posts

Re: Tensions in Google's ethical AI group increase as it sends demands to CEO

#411

Post by flatus »

throwaway2245 wrote: post #85
oh_sigh wrote: post #55

Earlier quoted context omitted.

Doesn't the obvious, honest answer make the most sense? Someone in Google management doesn't want their products acting unethically - merely because it is unethical?

This makes a lot of sense to me in the way you have said it. The department probably exists in the first place as a result of an internal political struggle and compromise - with some decision-makers wanting Google to be genuinely more ethical, and others wanting to avoid any genuine oversight. As a result, you have this bodge solution where the department gets to exist (helped by the PR benefit as mentioned elsewher…

throwaway2245:9 says >"The department probably exists in the first place as a result of an internal political struggle and compromise - with some decision-makers wanting Google to be genuinely more ethical, and others wanting to avoid any genuine oversight."I certainly hope you are right. And with some foresight, the ethics department is a "Golgafrinchan Ark Fleet Ship B" now fully loaded and ready for departure. Bon Voyage!

https://hitchhikers.fandom.com/wiki/Golgafrinchan_Ark_Fleet_...

Re: Tensions in Google's ethical AI group increase as it sends demands to CEO

#412

Post by pseudalopex »

ergocoder wrote: post #409
andrewprock wrote: post #356

Earlier quoted context omitted.

Sure, but you left out the actual situation: "If you eat dinner today, I'll quit my job" - then, you fire them and get on with dinner.

If I eat dinner, you will want to quit and I want you to quit. Both want to part way. We can agree to disagree on being fired or not being fired. But why would 'being fired' (or not) matter since both sides want to part way?

The debate about her being fired or resigning is a debate about who to believe about what led up to it. People who insist she resigned see her saying she was fired as evidence of a victim complex and publicity seeking. People who insist she was fired see Google saying she resigned as evidence of unfair treatment and dishonesty.

Re: Tensions in Google's ethical AI group increase as it sends demands to CEO

#413

Post by musicale »

JamesBarney wrote: post #307
throwaway201606 wrote: post #163

Earlier quoted context omitted.

I wish I had more upvotes to give for this comment. It gave me multiple upsights about how to clean up my approach thinking about the issues here: + framing an analogy using technology governance functions - whose functions and values can be clearly quantified and measured by many independently - with a ethical (moral) functions - whose values and results are much more nebulous + drawing a parallel between hackers -…

I read an interesting article that argued modern activists learned how to protest on college campuses which are unique due to the influence the student/activists wield. And that when these activists protest later in life they are not as prepared for the more challenging environment where your antagonist is unsympathetic to your cause. This seems to be playing out here. I can't really see how Google could acquiesce to…

Like many institutions, universities ignore all complaints through regular channels (and intentionally delay any consideration of action until after the troublemakers have graduated); the most effective leverage students seem to have is creating bad publicity, which affects the university's bottom line.

The lesson they are teaching their students is to be utterly cynical about going through the system, and to believe that creating a public scene or disturbance is the best way to convince organizations (or anyone else) to change their policies.

Students probably aren't being creative enough in this regard; there are lots of ways to achieve or threaten bad publicity, such as working with media.

Re: Tensions in Google's ethical AI group increase as it sends demands to CEO

#414

Post by dang »

Miraste wrote: post #378
dang wrote: post #376

Earlier quoted context omitted.

I'll edit it for you and detach this subthread. Thanks! I have this crazy idea to build a pull requests system for HN comments. People could suggest edits that way without cluttering the threads. It's either a terrible idea or a great one.

For users to PR their own comments, or for anyone to PR any comments? The former sounds overcomplicated, but the latter... I don't know how that would turn out, but it would be fascinating to watch.

I mean the latter. Yes, exactly.

Re: Tensions in Google's ethical AI group increase as it sends demands to CEO

#415

Post by thisiszilff »

GVIrish wrote: post #380
themaninthedark wrote: post #191

Earlier quoted context omitted.

Someone says they will quit noon tomorrow if you don't meet their demands. You will not meet their demands. You tell them not to bother coming in tomorrow and disable their security badge, did you fire them? They made the decision they did not want to work for you as had been previously agreed. You only changed the timetable. To put this another way, if I tell my company that I will be quitting in 2 weeks and they te…

So this gets into nuance and I will preface this with, we don't know what her email specifically said. If you give your company a date for the cessation of your employment, that is giving notice. If you say I "will resign if", is saying that you will do something in the unspecified future. So if you give your two weeks notice you have formally turned in your resignation. In your example I would agree that they just c…

From what I've gathered the contents of the email were largely like "Either we do XYZ or we can work towards setting a date where I leave the company." So it is open to interpretation and vague enough that both sides can get their interpretation in. The way I phrase it to myself is that she was fired for threatening to resign. IE, that relationship wasn't good to begin with and this was just the last straw that pushed them to severing it. I suspect that people who try to claim she was fired/that she resigned don't focus enough on the broader context where that relationship was bad in the first place. This doesn't seem like it was something out of the blue, this has definitely been going on for a bit, so given that, which party took it over the line to breaking up their employment relationship seems like a detail rather than an important fact.

Re: Tensions in Google's ethical AI group increase as it sends demands to CEO

#416

Post by thisiszilff »

pron wrote: post #260
lawnchair_larry wrote: post #243

Earlier quoted context omitted.

This comment (and these foolish entitled SJW employees) fail to observe that companies like Google only exist in America in no small part due to that fact. No other country has produced companies that are even remotely comparable, and the US produces them regularly. Assuming, of course, that one’s preferred alternative doesn’t involve serving the CCP. Maybe some folks prefer it that way, but they seem all too happy t…

> that companies like Google only exist in America in no small part due to that fact That's a non-sequitur. There are many things that make America, and every other country, unique. Maybe America produced Google because it's the only country with that population size that's made up virtually entirely of immigrants? It's also an is-ought fallacy. America is also the only rich country that's built that much of its weal…

> It's also an is-ought fallacy. America is also the only rich country that's built that much of its wealth on slave labour, so does that mean slave labour is a good thing?

I'm curious about this one. Most of what I've read indicates that modern American wealth was largely circumstantial due to european economies being obliterated during WW2. Do you have a source for how much slavery affected American wealth?

Re: Tensions in Google's ethical AI group increase as it sends demands to CEO

#417

Post by thisiszilff »

cbozeman wrote: post #235
frongpik wrote: post #189

A big reorg is coming to that ethics group.

I think the problem with this group is they don't understand what their function is, and Google can't really explain it to them. Their job is to navigate, any way they can, a path that can be perceived as ethical for Google to exploit AI to further shareholder value without having to also deal with extreme pushback from media, watchdog groups, etc. These people probably think Human Resources exists to help them.

I think the best case scenario for AI ethics is that those researchers work to develop processing & metrics to ensure AI systems are fair. The current model seems overwhelmingly adversarial where AI researches do X and then the ethics researches come in an indicate problems P, Q, R. I'm curious how much internal communication there is between the ethics department and the AI departments. Instead of doing research on how the datasets are biased or something of that sort, it seems more fruitful to help researches develop balanced and fair datasets. AI is going to develop yes or yes. The best case scenario is that AI ethics is a first class concern for those researchers and is factored into their development rather than being a case of "look at this gross, unethical behavior in this model!"

Re: Tensions in Google's ethical AI group increase as it sends demands to CEO

#418

Post by cbozeman »

ntsplnkv2 wrote: post #377
cbozeman wrote: post #285

Earlier quoted context omitted.

The problem, as I stated in another comment, is that these people don't really understand what their job is and that's causing this confusion. Their job is to build a bulwark for the company to guard it against negative press, and that will allow Google to create an AI that leverages all its data and currently existing infrastructure to increase shareholder value. They actually think their job is to prevent bad thing…

And Snowden's job was to keep the US government's spy systems working. So what? Just because it's their "job" doesn't make it right or not worthy of whistleblowing. If companies are doing bad things we want whistleblowers to tell us. I just don't understand your argument - "you can do bad things if it's your job." No. It doesn't work that way.

Yes, you can do bad things if its your job.

You can't do illegal things. Difference.

Re: Tensions in Google's ethical AI group increase as it sends demands to CEO

#419

Post by cbozeman »

pseudalopex wrote: post #402
cbozeman wrote: post #193

Earlier quoted context omitted.

From what I understand, they didn't fire her, they accepted her resignation? When you append, "Or I will leave" to a request and the company isn't interested in your request, you're resigning. You can dress it up in fancy language, but that's the bottom line. Or did I miss something about this story?

She proposed working on a date. Google terminated her employment immediately and said it was for misconduct.[1] [1] https://twitter.com/timnitGebru/status/1334364735446331392

She said, or Google said?

An any rate, this seems to me to be an excellent case study in ultimatums. Namely, don't give one to someone more powerful than you.

Re: Tensions in Google's ethical AI group increase as it sends demands to CEO

#420

Post by GVIrish »

thisiszilff wrote: post #415
GVIrish wrote: post #380

Earlier quoted context omitted.

So this gets into nuance and I will preface this with, we don't know what her email specifically said. If you give your company a date for the cessation of your employment, that is giving notice. If you say I "will resign if", is saying that you will do something in the unspecified future. So if you give your two weeks notice you have formally turned in your resignation. In your example I would agree that they just c…

From what I've gathered the contents of the email were largely like "Either we do XYZ or we can work towards setting a date where I leave the company." So it is open to interpretation and vague enough that both sides can get their interpretation in. The way I phrase it to myself is that she was fired for threatening to resign. IE, that relationship wasn't good to begin with and this was just the last straw that pushe…

I'd agree that the relationship looked like it was all but unsalvageable. The thing I don't understand is why Google insists she resigned rather than just owning that the ended things early.

Saying she resigned sounds better I guess and is at least plausible to enough people that maybe they thought it better to characterize it that way.

But like you said, it's not really the crux of the whole situation. At least not IMO.

Post reply on HN