Comment by shoo | Hacker News Reader

Comment by shoo | Hacker News Reader

One thing I've wondered: what is the purpose of Google's ethical AI group? What utility does it provide to the corporation?

I suspect the primary motivation for funding ethical AI work is _public relations_ (or less kindly, corporate propaganda) - to help mitigate damage to Google's brands and reputation by public concern surrounding artificial intelligence.

It is well understood that it can be difficult to get someone to understand something when their paycheck depends upon them not understanding it. Conversely, it may be difficult to remain employed and continue drawing that paycheck if you publish research that does not portray your employer in a good light, especially if the underlying reason for your role is public relations.

Here's a fun experiment: open up Alphabet's annual report from last year [goog10-k2019], search for the keyword "ethic", and there are two hits, both under item 1A, risk factors. The latter keyword match for "ethic" in the 10-K is regarding supply chain risks, and the potential for unethical behaviour by alphabet's suppliers. But the former keyword match is far more interesting:

> Our ongoing investment in new businesses, products, services, and technologies is inherently risky, and could disrupt our current operations and harm our financial condition and operating results.

[...]

> In addition, new and evolving products and services, including those that use artificial intelligence and machine learning, raise ethical, technological, legal, regulatory, and other challenges, which may negatively affect our brands and demand for our products and services. Because all of these new ventures are inherently risky, no assurance can be given that such strategies and offerings will be successful and will not harm our reputation, financial condition, and operating results.

[goog10-k2019] https://www.sec.gov/Archives/edgar/data/1652044/000165204420...

Without actually knowing the history, I think it's probably stupidly conceived. We seem to generally imagine that the biggest problem or challenge with AI is ethics because we frame ethics as being largely rooted in something emotional, like empathy or compassion, and not logic.

The movie I, Robot explores what happens when "logic" is taken to its logical outcome by an AI. It posits a robot with feelings as the antidote to this issue.

We tend to overlook the fact that humans themselves come in varying flavors with regards to degree of emotional affect, empathy, compassion, etc and how ethical they are. We haven't gotten this sorted for humanity yet and when tech goes sideways and hurts many people, we like to blame it on the tech itself fundamentally lacking "a heart" or ethics instead of blaming it on the character defects of the architects of the tech.

We still haven't really figured out how to do human intelligence well and that's scary and not something we want to admit, so we project our deficiencies onto the tech and this incident is uncomfortably casting light on the fact that it's not really clear cut like that. It's really not the case that you can readily separate the people from the tech when it comes to things we label "ethics."

/Hot take

I love this take but it's way too over-thought. As is the PR explanation.

I'd be willing to bet some money on it being simply the result of some employees saying, "AI ethics is super important. We feel strongly there should be people working on AI ethics to make sure it's not racist or sexist". And anyone who objected was yelled at with words to the effect of, "are you saying ethics isn't important? do you not care about ethics?" until opposition was silenced and management felt it was a small sop to quiet the ranks.

The right approach to this would have been to say to those advocating for it the following:

"Ethics is of course tremendously important. That's why we carefully select our workforce and insist on the highest ethical standards from everyone, in their work. Ethics is not something that can be compartmentalised or delegated to a separate team. Ethics underlies all we do. If you see something you believe might be unethical, please raise it with your manager so it can be resolved company wide. Until then, return to work."

Creating a dedicated ethics team was a dumb idea because it can only possibly end like this: you hire people and tell them "your job is to ensure we are ethical" and they're not going to sit back and say, you know boss, I sniffed around and everything seemed above board. Nothing for me to do here right now. I'll check back in a month. Of course they're going to see unethical stuff everywhere they look, regardless of merit. Their salary literally depends on it.

I'd be willing to bet some money on it being simply the result of some employees saying, "AI ethics is super important. We feel strongly there should be people working on AI ethics to make sure it's not racist or sexist".

Oh, no doubt. But that scenario would be rooted in this larger context of how we generally relate to this topic.

Creating a dedicated ethics team was a dumb idea because it can only possibly end like this

I agree, which is why I say "it's probably stupidly conceived."

I find social stuff fascinating and I've had the following college classes: Intro to Psychology; Social Psychology; Negotiation and Conflict Management. I also raised two fairly Aspie-ish sons and I've had to explain social things to them in very exacting terms, which forced me to really up my game on being precise in order for it to be useful to them. My sons and I still look for things that meet our rigorous standards for "How do you know that's actually accurate?" because social phenomenon are inherently hard to test: People behave differently when they know "This is a test. This is only a test." than they do in real life (which is why I think UBI experiments are useless: These people know the study will end and the money will stop rolling in, so it doesn't tell you how people would really behave if we really instituted a UBI).

Anyway, I find social stuff fascinating and people routinely interpret my social observations as being ethical or moral judgements and I get this reputation for being a real judgy bitch, basically, even though I"m a lot less judgy than most people. But when push comes to shove, "ethics" is generally about determining which behaviors protect one or both of two things: individuals from group abuse or groups from individual abuse ("tragedy of the commons" type stuff et al).

So when I talk about what works or what doesn't, in some sense I am talking about ethics, but I'm usually not trying to be some kind of moralizing priss, which is how that gets interpreted. I'm usually trying to say something like "Objectively speaking, you can't be both heavy and light at the same time. That fundamentally doesn't parse." and then someone is all up in arms about "Don't talk about my momma!!!! grrrrr!" type stuff.

So I feel strongly that what they really were probably wanting to study is "How do we do AI well in ways that interact with this social stuff like racism?" and out popped the word ethics because that's how people think of that. And it was a terrible idea on the face of it because of the inherent assumption of judgment rather than a more objective framing about studying AI and social phenomenon and how does that work? When you start using AI to do things like parse faces, what are the practical pain points there that we need to be aware of and try to address?

But I think most people can't readily separate that out in that way, which is part of why I am so often in hot water for saying "Well, actually, I think that simply doesn't work because...." And when I say that about social phenomenon, I am often de facto talking about "human ethics" whether I want to be or not and it's very often a giant pile of manure that I've just stepped into.

Wow. That is a masterful explanation, and I think right on the money. Thank you. :)

Almost there. Each individual comes with a distinct set of biases. Social collective action is the contextual weighted average of those individual biases. This results in quite resilient societies: at equilibrium individual biases balance each other across the population, small changes in a relatively small number of individual preferences result in societal course corrections, avoiding dangerous shocks. Smooth ride.

A subtle AI risk is simply the centralization of the decision making process. This leads to the reinforcement of some subset of biases, until they completely overshadow alternative options. The system becomes rigid, having lost its day-to-day individual level feedback, thus its capacity to adapt to new circumstances. Sooner or later, it will fail.

See also 'too big to fail' and 'central planning'.

There's one other use for it - and maybe this qualifies as PR - hiring.

Super-intelligent folks are the target of google's hiring.

I suspect these types of people can work anywhere, they get to pick and choose.

If they can pick and choose, they will have other motivations, like being able to make a difference, and not blindly create a dystopia for everyone on the planet.

So... having an Ethical AI group lets people join google, look themselves in the mirror every day, and know their employer is trying to do things right.

I don't know what the reality is of an Ethical AI group. But having a good one that works might make the difference hiring exceptional people.

What makes you think those "super-intelligent" people would all agree what "do things right" means? What makes you think they would side with the Ethical AI group at Google? As an example, John von Neumann was as "super-intelligent" as they come, and yet he made it very clear he wanted to nuke Soviet Union.

could it also be possible an ethical AI group might provide guidance on what to do and where the lines are?

(and of course I don't know reality - for all I know enron had an ethics group too)

A less synical conclusion would be that the longterm viability of a AI-centric company like Google benefits a lot of reasonable development in the early stages. If Google can avoid ethical hiccups it might provoke less regulation and provoke it later.

All else equal, being a reasonable and ethical company IS good business. Problems arise with ethics and business collide.

This kind of liability mitigation use case is at odds with academic freedom. Your view would imply making AI part of your InfoSec / CyberOps team, not an academic research group.

the primary motivation for funding ethical AI work is _public relations_

Exactly. The secondary motivation is the possibility of benefiting from some basic research. It's ultimately a credibility play, which makes their handling of the whole affair so mind-bogglingly dumb. There would have been no problem at all for Gebru to have presented her supposedly anti-Google paper; zero practical consequence for Google.

On the contrary: The research paper is evidence of knowingly committing later wrongdoings.

As we debate this, Google has been credibibly accused of the most basic and wide-ranging antitrust violations, and whether they actually pay a price for it, assuming the accusations are true, isn't a slam dunk.

Knowingly committing ethical violations with AI seems likely pretty low on their list of concerns.

It is, but to the managers in charge of this part of the organization, they probably want to keep their options and not be cornered by bad PR.

It's just marketing/pr. It's supposed to give them a clean enough image while they keep doing whatever they want under the hood. It won't certainly alter the status quo or influence any decision that could result in reduced profit for shareholders.

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 elsewhere), but not to act independently. Anyone in the department offering genuine oversight is at risk of being forced out.

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_...

In a world rife with deceptive marketing and advertising, PR spin, greenwashing, lobbying, astroturfing, and backroom deals, taking a corporation's actions or words at face value without considering if there might instead be an ulterior motive rooted in avarice is simply naive.

In a world of lies, deception, half-truths, fabrications, and hypocrisy, it would be simply naive to put any faith into a random internet comment. What could your ulterior motives be for posting this?

Beyond PR, if all the ethical AI people work for Google and none for governments or other regulatory entities, then Google has cleared the path for themselves in terms of regulation or they will be able to run circles around whatever does come up.

Reply on news.ycombinator.com