Amazon shareholders reject facial recognition ban | Hacker News Reader

Amazon shareholders reject facial recognition ban

reuters.com

291 pointspetethomas285 comments

In my mind selling facial recognition seems no different than selling a hammer. It’s amoral, it can be used for good or bad.

Id much rather see laws to limit when and how the technology can be applied.

same thing with land mines? What about an autonomous drone?

Hell, even a bottle of aspirin filled with poison tablets secretly placed on a store shelf could kill a future serial killer.

I believe your judgement mechanism is meaningless. Even the most evil stuff can be used for good, and everything is a tool.

A better metric would apply one of the standard ethical frameworks, like utilitarianism or deontological ethics. (Is this technology going to provide a net benefit? / If everyone used this technology would we live in a desirable world?)

Or nukes?

I will accept your metaphor as valid if you can present to me a moral use of land mines.

I agree with you about autonomous drones. They have many moral uses. Weaponized ones, not so much.

That depends on your mortality. I see it morally right to use it on my enemy.

How does one use them on one's enemy without also using them against the long-term population in their affected area?

By ensuring that they have an automatic deactivation.

https://www.reuters.com/article/us-usa-landmines/u-s-halts-u...

"While retaining stockpiles, the United States has not used persistent mines for some time. The last field of persistent mines under U.S. control was in Guantanamo, Cuba, and was removed in 1999, the State Department official said."

How about enforcing a DMZ?

Yes, same thing with all of that.

A land mine can only be a landmine.

A hammer can build and destroy.

So it makes sense to ban mines completely, as there is no humane use case.

But for facial recognition, I do see valid use cases. So yes, I do agree with OP and rather have the bad use cases banned/regulated.

And your metric: "If everyone used this technology would we live in a desirable world?"

Totally depends on how it is used. I do see a potential world with facial recognition software and not be a dystopia. I mean, is all Sci-Fi now considered Dystopia?

How can for example robots really interact with humans without that technology?

A landline could prevent a burglar from walking on my property.

Or it could also kill a pizza delivery boy

Or yourself or the people helping you, after you had a stroke and don't remember anymore that and where your placed them.

Right, just like a hammer.

Landmines are a non-trivial factor in South Korea's freedom. For a large portion of their existence the South had an inferior army and economy. Their current might is a more recent phenomena.

Without doubt, landmines are a effective and cheap killing tool.

But they kill or cripple anyone. And the person placing them, usually does not remove them anymore.. so lots of one legged children in the end, as normally (rightfully!) war places see peace again. But with mines being cheap to deploy and removing them beeing very expensive they tend to stay where they are, until your childrens children walk over them as boyscouts.

And south corea had US backing as far as I know.

At the time the landmines were placed the US was at a stalemate with North Korea (and China) - and even the stalemate was pure luck as NATO forces were close to full defeat at many points after China entered the war. US backing did not make their survival a sure thing.

Another interpretation is, that the mines cemented that border.

Yes it did. But the alternative border would have been a unified Korea with the north’s leadership in charge. Whether that is worth the trade off depends on your politics I suppose.

Or a tight border for some years and than a reunion like in my country, germany?

(even though there were mines on that border, too, but in a limited way that there are no more traces of them today)

Anyway, I am still highly sceptical to the claim, that mines prevented a total communist korea. Debating about that is not really constructive. Do you have studies for your claim?

There's undeniably a freedom tradeoff here as facial recognition systems expand and many corporations and governments gain access to millions of capable sensors sending this data.

I can't think of a single use case where doing this in the cloud is a worthwhile tradeoff. If you want to do this locally, sure, I can see many silly use cases for it, but as soon as you start shipping that data off to say... Amazon, things get pretty sketchy pretty fast in my book. So while I agree there are good use cases for facial recognition, I feel these should never be implemented on centralized services.

Even locally is bad. Retail will use this right away to track everyone and what products they look at.

There are very, very few moral applications of facial recognition. Logging into your phone/laptop or assisting people with prosopagnosia, I guess. Everything else basically enables mass surveillance.

It is frankly astonishing, and an exemplar of technologist naivete, that so much effort was (and continues to be) dumped into developing this technology.

I googled it.

pros·o·pag·no·sia

/ˌpräsəpaɡˈnōZH(ē)ə/

noun PSYCHIATRY

a neurological condition characterized by the inability to recognize the faces of familiar people.

"The Man Who Mistook His Wife for a Hat" by neurologist Oliver Sacks is a fascinating story about visual agnosia - it was also my first introduction to artificial intelligence.

That's interesting, my Intro to AI professor also had us read this in college (UCI). Is this common introductory material for the field?

That whole book is fascinating!

Flagging “America’s most wanted” in public places like train stations or airports. Looking for missing children. Having home automation react differently to different family members.

Those are all usages I am personally totally ok with.

>Flagging “America’s most wanted” in public places

And what happens if we descend into military rule or a dictatorship? With "America's most wanted" fliers, the general public can just refuse to participate. With facial recognition, anyone deemed an "enemy of the state" can easily be tracked and eliminated.

Always assume the worst and work your way back to the good when doing the cost-benefit analysis.

Laws have a tendency not to stop dictators. They are the law.

In reality laws are only as powerful as their enforcement, and if there is no will to enforce, like what happens during transitions to dictatorships, then they’ll just be ignored.

Yes, but that is kind of the point. The lack of checks and balances tend to lead to totalitarianism.

So in what way will facial recognition and "criminal spotted on 5th avenue" either dissuade or encourage totalitarianism/removing checks and balances?. If a Gov't falls into totalitarianism, a facial recognition ban (both internally within AMZN as well as on a national/local level) won't stop them from requiring AMZN to provide them facial recognition behind-the-scenes.

I agree with the general point that it isn't a good method of stopping a government falling into totalitarianism. However it is helpful if there is general consensus against facial recognition because it makes it a bit easier to identify the authoritarian-leading parts of government because they can't hide behind 'it is just usual practice'.

There are other, better, arguments in favor of personal liberty to use against facial recognition by law enforcement. Efficient, highly automated systems crush people who just happen to get caught up in them.

Hypothetically?

A tool that could track individuals and was under the perview of Executive Branch (eg Homeland Security) sounds pretty dangerous. Any political rivals of family members who go for STI screening, abortion clinics, gay bars, oncologist, or couples consoling could be either leveraged or leaked for political effect. Or you it could be used to track groups of people, activist, reporters. Or any number of ways I wouldn’t be able to think of.

What I was trying to say is that if you are trying to establish a totalitarism not having any checks and balances on something like that would be handly.

> So in what way will facial recognition and "criminal spotted on 5th avenue" either dissuade or encourage totalitarianism/removing checks and balances?. If a Gov't falls into totalitarianism, a facial recognition ban (both internally within AMZN as well as on a national/local level) won't stop them from requiring AMZN to provide them facial recognition behind-the-scenes.

Technologies take time to develop and deploy, and the fact of their development and deployment is a red flag that tells you something bad is coming and gives you time to react before it can actually be put into place.

Setting everything up for turnkey totalitarianism is not wise.

And dictators have a tendency to use expensive infrastructure that was built before they came into power. Setting everything up so that switching from a free society to a mass surveillance and state control hellhole is as easy as editing a config file is not a good idea.

Yet it will still be as easy as deploying a docker container from the repo and hooking up the API.

I think what's more important about not building surveillance infrastructure is not that the infrastructure does not exist, but that building it does not become normalized. It should be morally and socially unacceptable, so those hired to build and operate one will either refuse or sabotage it.

Some people consider violating laws immoral per se. If you make actions leading to a dictatorship illegal, you reduce the potential dictator's supporter base.

Also, in a functioning state, the bureaucracy (mostly) follows the law and the judicary's interpretation of the law. Then again, functioning states don't tend to turn into dictatorships.

(Fun fact: Article 20 (4) of the German constitution explicitly authorizes every German to resist when someone abolishes freedom and democracy. Considering the mentality of the average German, this is probably necessary.)

Will the dictator obey the rule about facial recognition?

If the technology doesn't exist or at least isn't already installed, it's significantly more difficult for them to abuse it by just ignoring rules.

> the general public can just refuse to participate

The general public will be misled by the media. Already politicians are claiming that the media isn't truly independent or objective with claims of "fake news". If the media falsely claim that you are a dangerous criminal, the public will eat it up and ask for dessert.

> And what happens if we descend into military rule or a dictatorship?

We wear masks.

A guy in england was just cited for covering his face.

https://news.ycombinator.com/item?id=19928231

While I agree with the aspiration, I don’t see how it’s possible to use facial recognition to flag “america’s Most wanted” or to look for missing children without mass surveillance though?

It only works if it scans _everyone_ .

Additionally what happens when the technology isn’t perfect and innocent people get mistakenly flagged as persons of interest?

The other thing is once it’s installed and in operation, what’s to stop it being used for other purposes? - being used to target people peacefully protesting against the government or whatever.

It’s a slippery path from there into a surveillance state - China is already pretty much there with their social credit system - https://en.m.wikipedia.org/wiki/Social_Credit_System

I for one don’t want to wake up someday soon and discover it’s 1984...

And that’s the argument the grand-parent is making. It’s a tool. It can be used for good or bad. There are other tools like that.

Because of that, there’s unlikely to be universal acceptance or rejection. And without popular opinion, it will be hard to pass any laws that change the status quo.

CCTV cameras are already ubiquitous. That ship has already sailed. And honestly, there was never an expectation of privacy in public in the first place.

CCTV cameras monitored by humans are COMPLETELY different from a facial recognition system recording the identities and movements of all people. There is no comparison.

There absolutely is an expectation of privacy in public. Being seen in public by a series of uncoordinated people is massively different from a PI tailing you and recording your actions. This form of privacy is generally termed "obscurity".

> Being seen in public by a series of uncoordinated people is massively different from a PI tailing you and recording your actions.

That is actually completely legal to do in all circumstances, precisely because there is no expectation of privacy in public.

> CCTV cameras monitored by humans are COMPLETELY different from a facial recognition system recording the identities and movements of all people. There is no comparison.

Which is not necessarily how facial recognition would necessarily work. More likely would be to scan for known suspects and fugitives. But then we’re back to the “how is it used” question.

It is legal for a PI to tail you not because of the lack of expectation of privacy in public, but because it is impractical to have PIs tail everyone all the time in public. People are generally okay with targeted surveillance. Mass surveillance is the issue. Quantity has a quality all of its own, as they say. There has been a court case which addresses this issue[0].

Any technology which searches for fugitives will necessarily scan everyone. There is no such thing as targeted facial recognition, it can only work by mass surveillance.

[0] https://www.nytimes.com/2019/04/17/opinion/data-privacy.html

You're assuming facial recognition would necessarily be the moral equivalent of "[having] PIs tail everyone all the time in public.". But for that to be the case, facial recognition would have to scan every face it sees, store that face, and then cross-reference every other face it sees against every face it has ever stored.

I'm suggesting a far simpler use case: It scans your face and if you don't match any of the fugitives it's looking for, it forgets about you. I think this use case is far more likely because it's a lot simpler and cheaper to pull off. And that's a big difference.

Agree. And to take it even further, consider Bostrom's "Vulnerable World Hypothesis" where technology could arise that makes it trivial for individual actors to cause mass destruction. From the abstract:

"A general ability to stabilize a vulnerable world would require greatly amplified capacities for preventive policing and global governance."

https://nickbostrom.com/papers/vulnerable.pdf

What happens if the system flags the wrong person as "America's most wanted" and a trigger-happy cop is nearby? Not saying all applications are useless, but most of them definitely warrant additional thought.

6’0” black male with athletic shoes.

And we're already into "Think of the children!"

Those are the 𝘄𝗼𝗿𝘀𝘁 possible uses for this.

To illustrate lets imagine a really really good system that had a 0% false negative rate¹ , and a 0.000001%² false positive rate. If we were to sample the entire country looking the the FBI's most wanted we would end up with ~3290 matches, 3280 of which are going to be other people³.

Considering high value of the individual the chances of harassment or wrongful arrest (or worst) is pretty high.

1: In reality the false negative and false positive rate are going to be directly inversely related. The more you decrease the false positive rate the higher your false negative rate is.

2: That is 1 in 1 million

3: In this ideal situation it is assuming there is an even distribution. At this point computer vision is significantly worst for false positives for people of color.

It's definitely a problem if you use the face recognition as your only criteria - but if it alerted security to pay more attention to someone or to double check their documents?

Even with the false positive rate it's much more likely an individual identified by the system is a target than a random individual.

> > To illustrate lets imagine a really really good system that had a 0% false negative rate¹ , and a 0.000001%² false positive rate. If we were to sample the entire country looking the the FBI's most wanted we would end up with ~3290 matches, 3280 of which are going to be other people³.

> It's definitely a problem if you use the face recognition as your only criteria - but if it alerted security to pay more attention to someone or to double check their documents?

Where do you imagine this is happening? If its in an environment where papers are expected as standard (e.g. an airport), sure, this is relatively benign. If it's when you're walking through a mall and now you're being stopped and searched, how is that not bordering on harassment?

> Even with the false positive rate it's much more likely an individual identified by the system is a target than a random individual.

Much more likely than without the system, sure. Chances of it being a random individual and not the target? 99.6% (using the math above of 3280 innocents to 3290 people identified).

If a system had a failure rate of 99.6% when harassing people, I don't think I'd describe that as 'much more likely' to be a target than an innocent random.

Wouldn't it be the same as a police officer having a board of wanted people on his desk and (mistakenly) thought a person in a mall to be one of those? He'd check their papers (sorry, I'm European) due to his hunch and decide?

I believe there is no better way to identify wanted criminals at scale, do we want to comb the streets with policemen and "harass" 10x more people in hoes of finding them?

It's not a death penalty to be checked and bothering additional 3200 people in the whole country is a ok tradeoff to find wanted person in my opinion.

It feels like you want to cripple the way police looks for criminals because they don't act very kind towards suspects. Maybe police is the problem and not surveillance?

The reality is that police mostly catch “wanted” criminals during traffic stops or at their homes, or at the homes of known associates. Or when they get arrested for other crimes. People have a way of turning up.

How would this face camera thing even work? Like, a camera identifies so-and-so and a police car rushes out immediately? That’s not really how police do stuff.

If a system identified individual has a 0.4% chance of being on the FBI most wanted, then that is 'much more likely' than an individual picked at random who has a likelihood of 1 in 30 million.

While valid, your criticism is with the actual ability/execution of the technology, not the concept. Imagine we get it to a 100% match or, more realistically, use it in a way that makes more sense than just flagging someone and ruining their life forever (not to mention being biased). Would you still be opposed to the technology? If so, your opposition has nothing to do with the above argument.

This may or may not be valid, but I'm not aware of any advanced technology that is 100% accurate that exists today.

Rather than the hypothetical of 'is this ok if we have 100% accuracy', how about 'is 100% accuracy ever likely to be possible?'

If it's not, then the previous question is about as relevant as 'if unicorns existed, would you want one?' (and yes, I would!)

I agree, which is why I wrote "or, realistically, use it in a way..." etc.

> or, more realistically, use it in a way that makes more sense than just flagging someone and ruining their life forever

I think that's less realistic than developing 100% accurate algorithms, but now we're into anthropology instead of technology...

> Would you still be opposed to the technology? If so, your opposition has nothing to do with the above argument.

That's kind of a denying the antecedent propositional fallacy.

I am not advocating for large scale facial recognition, but your argument holds true for basically any measure law enforcement has to apply to locate missing people.

How big is the false positive rate with tips received by phone? Law enforcement probably has to deal with much more wrong clues on a daily basis than to double check an image that was flagged by an automated system. And if even a human inspection can't tell the difference between the missing person and an image of the missing person then it's definitely worth checking out.

Your (USA) police forces are crooked. They swat people on a first signal they've received.

Instead they should attempt to verify first and try to approach people in civilised manner, not shoot their pets...

That's a minority of police forces. There are a lot of police that truly care about doing the right things. The media covers corrupt police commonly and caring police rarely.

But you don't see such news come out from UK, AU, NZ or most of europe.

This is still better than the current paradigm of LEOs stopping and frisking anyone who "matches the description". That phrase has basically turned into carte blanche to harass anyone they want with plausible deniability. With facial recognition at least now there's one more constraint they'd have to operate under for probably cause, right?

That's true only if this somehow replaces the current paradigm. New technology generally supplements existing methodologies rather than replacing it.

See "Minority Report". Also, it is always a question of who will watch the watchers. If "recognition" is deployed, it will be taken as proof and now it is a weapon not a shield.

It is a misconstrued quote now but Ben Franklin said "Those who would give up essential Liberty, to purchase a little temporary Safety, deserve neither Liberty nor Safety."

Some of the developers may be naive, as people often willfully are for the preservation of their paychecks. The governments and corporations funding the development know exactly what they're doing.

> There are very, very few moral applications of facial recognition.

I mean, we don't even have the technology yet, we don't have almost any real entrepreneurship around how to use it... are you sure there are no good, moral use cases to discover?

I'm afraid that this position is as naive as the flip side - techno optimists who assumed that an 'open internet' is an ultimate good, and there are no bad uses for it. There's definitely more nuance than "it's 95% bad".

There are lots of applications. For instance, I would love to have an app that gives me contextual information about the people I am talking with. Specifically about the things that he/she already told me, as their name. I don't see anything inmoral un this.

It’s entirely possible I’m being naive about this I think that’s a fair criticism. I do also think that it’s equally naive to attempt to block any single company from developing this technology. As long as there is profits to be made company’s will develop facial recognition technology. This is why I mentioned regulating it’s use.

In general, once the proverbial genie is out of the bottle I don’t think you can put it back in in today’s world.

Sure, anything can be used for good or evil, uranium, an ICBM, whatever.

And yet you can't buy hand grenades at 7-11 even though that would be insanely useful.

No. PRECAUTIONARY PRINCIPLE http://fooledbyrandomness.com/pp2.pdf

The issue is not whether it is amoral or not to do something. The issue is in the inherent asymmetry in making an error.

If i knowingly sell a hammer to a particularly wicked individual, perhaps a dozen or so innocent people will perish due to my actions.

However, if i sell facial recognition to the wrong individual, I may have wiped out an entire generation, gene pool, race, etc.

The probabilities of either event are irrelevant. What matters is the consequences of the fat tails.

Wicked use of a hammer != wicked use of mass surveillance.

What about the possibility that a bad person uses the hammer to build a sarin gas bomb, uses it to destabilize the government, seizes control, and then executes entire races and gene pools? It's all about the consequences, not the probability, so doesn't that make selling a hammer just as wicked?

The US government routinely monitors purchase and search histories to prevent the active ingredients of something destructive like that from falling into nefarious hands, and you're posing a wholly disingenuous argument

Gp isn't remotely disingenuous, they are pointing out that 'probabilities ... are irrelevant [w]hat matters is the consequences' is a philosophy guaranteed to lead to stupid decisions.

To make a sensible decision in a risky context both probability and consequence are needed. There is always a (practically 0%) risk of any action leading to catastrophic global consequences.

In addition, the original Precautionary Principle paper linked is an idea that sounds nice but is actually a bit dangerous. If we applied the PP to the formation of the internet it should have been suppressed right from the get go - we never had and still have no positive evidence that social media won't cause extreme harm to the public (eg, it could have caused mass depression and disengagement with reality and maybe moral corruption :p). The internet is risky - did huge damage to existing industries and triggered the rapid rise of China through knowledge sharing and enabling remote management of industry (why else have we never seen anything like it before?). Impact of research into splitting the atom - did a lot of damage, much more risk.

Proving that something is safe without just doing it is hugely expensive. Technological advancement would not be a fraction of what we have now - the cost of trying things needs to be cheap to encourage progress. We couldn't prove most of the things we do are safe without being able to point to them and say 'seems to be working?'.

Not only that, the precautionary principle applies equally to inaction.

Should we research nuclear physics? The worst that could happen is global thermonuclear war, so forget about it. But what's the worst that could happen if we don't? We might need a nuke to divert an asteroid. We might need nuclear power to prevent climate collapse. It could save the world rather than destroy it.

"What you don't know can't hurt you" is contrary to evidence.

In that case, I would definitely like to see some facial recognition technology so its easier to catch that guy.

If you were going to commit ethnic genocide, how much help would facial recognition software be? Even birth certificates would be more helpful, if the problem was identifying your victims in the first place.

It's more like selling a sword. You may use it to chop up watermelon but we both know that's not the main use case.

All technology has a moral component, it's never neutral.

I still don't know if I fully agree with Postman's Technolopy thesis, but he definitely challenged my younger technophilic self to be more critical.

https://en.wikipedia.org/wiki/Technopoly

Technopoly: The Surrender of Culture to Technology is a book by Neil Postman published in 1992 that describes the development and characteristics of a "technopoly". He defines a technopoly as a society in which technology is deified, meaning “the culture seeks its authorisation in technology, finds its satisfactions in technology, and takes its orders from technology”. It is characterised by a surplus of information generated by technology, which technological tools are in turn employed to cope with, in order to provide direction and purpose for society and individuals.[1]

Postman considers technopoly to be the most recent of three kinds of cultures distinguished by shifts in their attitude towards technology – tool-using cultures, technocracies, and technopolies. Each, he says, is produced by the emergence of new technologies that "compete with old ones…mostly for dominance of their worldviews".

If there is a ban, it should be country law, not just one company. How is this gonna achieve anything if another company goes ahead anyways?

This same argument could be made about any decision to abstain from anything. If a country banned facial recognition the work could happen in another country. You can make it even more farcical by changing the act you abstain from: Why not kill someone? They will die eventually.

For any actor, their choice to do or not do something matters. It might not matter much, but it has some impact. I would not describe Amazon as a small player or having a small impact. Dismissing the importance of what they choose seems myopic. To return to your frame, I'd say that if Amazon stuck to its decision not to work on the technology it would be more meaningful than any national ban outside of, probably, the US or China.

In general - it seems like, if we don't like something, we should celebrate it not being done. Even if it's not done in some limited context. In fact, outside of a world government with effective law enforcement - all choices not to do things are limited. So it's really all we have!

Yes, AI regulation needs to be global to be effective.

The difference between "we should abstain from our moral values because everyone dies anyway in the end" and "we should find ways to pursue our moral values effectively" is substantial. You don't have to be nihilistic to not accept cheap signaling being sold as a solution to real problems.

Few corporations deciding to not do dirty work, just providing their lower-level cloud resources to thousands of smaller firms who will gladly compete for any kind of money -- that does not solve anything.

Stuff like GDPR can change at least something.

> AI regulation needs to be global to be effective. [...] not accept cheap signaling being sold as a solution to real problems.

Amazon making a pledge not to work on facial recognition does not seem like cheap signaling to me! Cheap signaling would be Google's "AI Ethics Board" staffed with people unlikely to make trouble while Google continues to work on AI. A pledge not to work on technology is an expensive signal!

Ideally, sure, to stop something everyone agrees not to do it. But history is stuffed full of examples where one company could have made a difference. Imagine if IBM refused to work with the Nazis or if Dow Chemical refused to sell agent orange to the US.

It is impossible to solve anything "forever." Facial recognition will be developed, but how long that takes and how well the world can grapple with the consequences matters!

At the time there was no alternative to IBM and their technology. Same can be said of Dow Chemicals.

Now there is competitors. Alibaba Cloud and Tencent Cloud will gladly help anyone with facial recognition . See how evil America refuse to help you with your needs? We will because we care about your needs, China will support you.

That's incorrect.

There were other tabulating machines (for example the Bull[1]) and of course there were (and are) other deforesting agents. The governments in both cases would have found another supplier.

It also would have made a difference! There's a reason the Nazis picked IBM and the US picked Dow.

If it's the case that Amazon could do better work than Tencent and Alibaba - then their abstention is laudable. If it's the case that Amazon is inferior, then they'd get out-competed in a winner-take-all marketplace so they both save money and have clean hands. So win-win.

[1] http://www.feb-patrimoine.com/projet/bull_t30/tabu_t30.htm

You've basically described the prisoner's dilemma. However, there are stable strategies where cooperation benefit all parties.

Would you care to elaborate on how these might be deployed to form policy on this particular issue?

The point is that if enough ethically-minded people who believe facial recognition is bad see Amazon doing things like facial recognition the company will lose customers and staff. The people proposing a ban presumably feel there enough of those people now (and in the future) that Amazon's bottom line is under threat and it needs protecting from itself.

The idea of outsourcing your company's ethics to the government seems very shortsighted to me. Facial recognition may never be banned, yet privacy tech could become a strong factor in where developers want to work. Consequently Amazon are right to consider this sort of thing long before the government steps in.

I have probably a Pi radians stance from you. I don't believe Amazon (or Facebook or Google or ATT) will necessarily lose a meaningful number of customers. I don't think people are that cohesive in their thinking. It seems to be too easy to roll out benefits to the customer so they don't care about some abstract downside.

So if Amazon plays unethically, and plays well, and you are competing, you are going to be seriously hampered by not playing along too. And you will have a hard time justifying it to your shareholders or potential investors. Amazon couldn't even win this argument and the game's just starting.

I think ethical companies like the kinds of regulations that keep the playing field level and make ethical transgressions illegal. It gives them a strong support to make their ethical stand.

Look at all of the information that's being gathered today with no respect for the individual's privacy. I'm old enough that this is outrageous, and I really can't believe that people are standing for it. Yet it seems that every new app builder has to enter the market with a data-collecting free offering. If this were illegal, then they could compete on quality and price.

One thing I've learned as I've gone through some things is that if you're the only one in the game who's playing by the rules, you're going to lose.

One simple ethical rule that I live by:

If I ever find myself saying “if I don’t do this, somebody else will”, I don’t do the thing. It’s a good indicator that the thing in question is the wrong thing to do.

Other people will still do it. That’s true. There will always be evil people in this world. But it won’t be me. And that’s important.

uhh wouldn't that mean you don't do anything? Not everything you don't do is evil ;)

A very noble approach, however in the life of business, moral has no place. Which is very unfortunate in my opinion. I couldn't imagine how the world would look like if everyone thought the way you do...

Doing something legal but immoral to boost your profit margins is by definition immoral. I don't know how some ceos sleep at night.

People who don't sleep at night when they do immoral things don't tend to become CEOs.

Probably on stacks of money.

In capitalism system everything is equated to money: ethics, success, self-worth.

To be exact, it is tied to short term earnings.

I remember a documentary about drug dealers, and these were the exact words that they used to justify their deeds: "If I don't do it, someone else will, so it makes no difference".

That's a good and reasonable personal ethic, but it transfers into a (imo) poor, even disingenuous public discourse.

Let's say we're talking about the clickbait/trolling problem in political discourse. We can berate some prominent individual, but that individual is inevitable. If she wasn't doing the clickbait, someone else would be and we'd be beratying them.

Focusing on the individual when a problem is systemic can be empty, righteous moralising.

The individuals come together to create a society with rules. To be a part of society one must follow those rules.

Ethics, to me, have a place there, and it would be better in all cases to establish those rules or norms in your society, or your culture, rather than regressive actions like banning something.

Bans are a type of rule/norm.

Is case by case, in this case if they think is so bad, say selling hard drugs, should be a ban. Facial recognition banned is like AI banned under the same reasonings. Others will be doing it.

Because the USA is failing to act on climate change (and join Paris), neither should the rest of the world? Then we can certainly abandon hope right now.

At the same time China is producing more banned CFC, no one is doing anything.

Yeah, I never got this American Disdain for the State and regulations. Everything - health, pension, parental benefits, environmental restraint, repression of discrimination - ends up in the domain of corporate benevolence, the whims of the market and of shareholders.

If it’s not in the Law, it's just corporate marketing

It exists because the American government (or governments writ large around the world anyway) have had a poor track record of regulating things ethically. For example, the us government has not done a great job of ethically regulating native Americans, or, ethically regulating terrorism abroad, or maybe in more quotidian and economic terms, regulating the sale of loose cigarettes. It baffles the mind that one would want the same entity responsible for such large scale horrors responsible for deciding educational policy.

If you don't want to involve regulations resulting in people dying, we can look at regulating interior decorating, or hair braiding, neither of which was conducted in the public interest. Can you name the last 2000 regulations that have been enacted in the us? Can you be sure all of them were in the public interest?

Regardless of how well the democracy your government is based on is functioning, I have a rather hard time imagining that a corporation focused on maximizing shareholder profit would do anything better than a piss-poor job for the vast majority of the population.

there's a wide amount of space between 'regulation by government' and 'regulation by corporations'

Option A: Government regulation, sometimes fails(Article 7 and 11), sometimes get things right(GDPR).

Option B: Corporate entity self-regulates in favor of profit. consumer is screwed.

Corporations regulate not for public interest, but for profit, sometimes it happens that those two are aligned - but that is an exception - not a norm.

If someone wants to pay you to do something immoral, then that means that you're the best/cheapest/some mix of that they found. If you refuse, you make an immoral act more expensive or reduce its "benefits".

That can be enough to reduce the cost/benefit ratio far enough that your potential customer does not consider it worth it anymore. If it does not, the next best option also has the option to refuse, and so on. Even if the immoral act eventually gets done, you have harmed the immoral actor, thus reducing their ability to act in the future -- and it's likely that you wouldn't approve of their future actions either.

Because when another company fills the void they'll know what happens if you don't think about these issues and risk getting shut down themselves.

That opens up the possibility of benevolent actors filling the void. Or, someone at least as bad but better at playing the government.

Doing the right thing can be excellent PR and boost customer acquisition and retention - especially when you are consumer facing.

What it would have achieved is to enlist Amazon as an ally in lobbying for the necessary legislation.

Corporations don't like to watch their competitors profit from things they are not allowed to do themselves.

Love that the article says people can use the AWS support system to report issues with the technology. That's a black hole of no return if you are a paying customer much less someone random off the internet.

Let me just make a distinction that if you are a support paying customer, you get very fast and somewhat (depends who you get) helpful support, including real time webchat or through the phone.

It's arguable this should be included at all levels but I guess AWS wants to keep their support top notch for the larger pocket users.

The one time I raised a ticket on a corp AWS account, it was handled swiftly and competently. I'm sure it depends much on what your issue is.

Every ticket I've ever created on the paid Developer support plan was within the "24 hour" and "12 hour" windows for guidance and system impaired issues. However, always always always at the far end of those windows.

Meaning if I had something I really needed help with, it was going to be a conversation for the following day.

I'm not really mad, we're just pre-release, and the number crunching to go from Developer adding 3% to cost vs. Business adding 10% to cost is too damn high. Our infra is about $10K/mo right now (pre-release), so we're talking $1000/mo for live chat and a shorter window vs. $250-300/mo we're paying right now.

Can one report an AWS issue from a TSA inspection room?

No, you can call your lawyer instead.

Even lawyers have a hard time dealing with secretive no-fly lists - I wonder if face triggering by resemblance to data on a list is accessible enough to contest.

I feel like there are clearly ways in which facial recognition can improve a lot of systems -- ticketing (think airports, sporting events, concerts etc); automated stores (like the amazon cashierless stores); payment (imagine not needing a wallet because the combination of your face, gait, and voice is enough to verify you); security (home/building/car entry, detecting suspicious individuals, known terrorists, past offenders, restraining orders)

There are also the many valid awful ways it which it can be used, particularly by the government. Why not just ask for laws preventing the bad uses? I don't get why it has to be all or nothing. If it's strictly government use that is concerning, or police use, then pass laws banning government collection of facial recognition models. There's no need to pass laws banning all forms of racial recognition simply because there are certain dangerous use cases.

It'd be like if back in the 20's or 30's someone said we need to ban cars because they could someday be driven by terrorists into large groups of people. It's a tragically valid concern, but we all agree it's not reason enough to ban cars altogether. An even more realistic example is folks back in the day complaining about how dangerous cars were, and rather than banning all cars deciding on speed limits and safety belts and airbags and anti-lock brakes and so on. It doesn't have to be all or nothing.

I would quibble with your basic premise of facial recognition improving these things or even if these things are worthy goals. At least from my personal and selfish interest I don't perceive any improvement offered for myself or society. From a perspective of a shareholder/interested-party looking to maximize X via Z then yeah it makes sense.

If it came with absolutely no downsides, of course it would be awesome if a subway could identify and charge you just by walking in. Or an Uber. Or a grocery store. It’s the perverse abuses of such technology by the government or advertising companies that makes it dystopian.

Honestly even if the tech were magically perfect I think the maximum benefit is pretty small. Pulling out my wallet to pay for something or show my id only takes a handful of seconds and it's a routine action with minimal cognitive load. Plus it has the benefit of me knowing exactly what's happening with the transaction. I can decide on the spot to use a different card for whatever reason and I know precisely when a transaction is taking place.

I'm sure the automatic charges described here would be loved by corporations, which would try to find ever more clever ways to manipulate people into making even easier impulse buys.

Indeed, convenience at all costs is a curse that's driving towards 1984. The destination of this absurdity is why even bother thinking if someone else can do it for you...

> security (home/building/car entry, detecting suspicious individuals, known terrorists, past offenders, restraining orders)

So now we literally destroy the lives of anyone who has ever committed a crime in the past. Imagine a world where every single store refuses you entry because you stole something years ago.

That’s the other thing that doesn’t get mentioned more in these discussion. Governments can be bad actors and suppress citizens like how China is doing right now, no guarantee your own country won’t use it to suppress citizens who they don’t agree with.

So we pass a law to prohibit that.

Passing laws like that is usually seen as "weak on crime" and is unpopular with politicians.

Because corporations and governments always follow the laws they write and laws never get repealed and overridden.

The fourth amendment already has a bunch of exceptions to it, and the constitution is the highest legal framework in the land. Other laws mean diddly.

Laws are a high latency side-chain of power and authority.

That doesn't mean they are completely useless.

I'm not sure anyone is proposing that laws are _entirely_ useless. But it a tech is powerful enough, any abuse at all has potential for widespread negative effects. In the US, we've already seen post-9/11 how laws are side-stepped, trampled or quietly ignored even by well-meaning actors. Plus, getting laws in place is _hard_. Much of the public doesn't care enough to make noise about this kind of tech until it has already been abused (and sometimes not even then). None of this is meant as an argument against your position per se; just counter-point.

Sounds like they'd be leaving a lot of money on the table. More likely these people would be able to shop at retailers that took that into account and took appropriate precautions. People refuse to do business with people for reasons that are much dumber or more cynical than "has previously ripped off a similar business."

If you haven't already, I'd highly recommend reading 1984. I'm sure you read it in high school or whatever, but actually read it as an adult.

Definitely a good idea. But the problem is that it's gone down the memory hole, https://www.nytimes.com/2009/07/18/technology/companies/18am... . It's a bit scary how much of it has been happening, and it's not been completely that either; there's a lot of elements from the other side of the picture. Typically espoused by Brave New World.

I wonder if the personality test Amazon has during the prelim-interview can detect if you've read such books in the past.

They'll just break the laws and pay the fine.

“payment (imagine not needing a wallet because the combination of your face, gait, and voice is enough to verify you)“

If it can be done from a distance, this is bone chilling to me. A company could use this to literally extract money from someone’s bank account.

I agree with the rest of your sentiment though.

Whats the big deal to not carry a wallet? It’s not the convenience I’d swap for privacy. If you imagine where this could go wrong it becomes easier to just stick to the wallet. Even to cash.

It seems to fall under the quiet assumption that technology advances and automation are inherently good, and that manual processes and analogue things are inefficient and undesirable.

We don't even do the whole 'should we' evaluation because it is assumed it will happen anyway somehow and we'll figure out the rules after the fact.

> I feel like there are clearly ways in which facial recognition can improve a lot of systems -- ticketing (think airports, sporting events, concerts etc); automated stores (like the amazon cashierless stores); payment (imagine not needing a wallet because the combination of your face, gait, and voice is enough to verify you); security (home/building/car entry, detecting suspicious individuals, known terrorists, past offenders, restraining orders)

I’m not even sure how facial recognition brings anything fundamentally different to the table in these examples: no airport will let you board only by facial recognition. Even payment is already super fast with contactless terminals, and credit card companies will never allow low fees billing by face recognition alone.

Every other low stake application already has a somewhat fast biometric option that didn’t take off because people just don’t want biometric databases of them (either by just being too lazy to opt in and give the info, either by clear opinion)

Cars are a device of personal freedom that you can own and operate yourself

Facial recognition tech is a device of surveillance owned by governments and large corporations that do not enable greater individual freedom, at best a few convenience benefits at worst total restriction

One of the things that blows my mind about the current tech landscape is that Microsoft now appears to occupy the moral high ground with its stance calling for regulation on facial recognition.

I grew up with a profound distaste for Microsoft's anti-competitive business culture under Gates. But now I have to give credit where it's due. Even Apple, which has come out with strong statements on privacy, is lagging behind MSFT in this area as far as I can tell.

The sad thing is that Microsoft really haven't improved. The competition got more successful and took the ethics down much lower than the bar set by Microsoft. They only look good because their power diminished and others are more effective at being utterly horrible.

Apple are but one example. Can you imagine if Microsoft had tried to take 30% of your sales revenue for selling software that runs on windows? Apple did it, got away with it and took that bar down lower. Google followed and got right in there. Privacy concerns? Race to the bottom with facebrick, goog and everyone else seeing who can fk everyone over with a bait and switch strategy, deeper, faster and more profoundly. Microsoft haven't really competed there but I'm pretty sure it's not because they've taken a moral position anymore than IBM did about competitive practises in the 90s.

Microsoft probably haven't got a lot of my data under false pretenses and without my explicit consent but it's unlikely because they chose that - they just don't have much I'd want to use as a lever to rip me off. At least yet.

IBM look ok now, right? Whodathunk they literally provided the infrastructure for Nazi death camps? [1] Did their investors bless that moral position in those days too? There go Amazon claiming the right to sell facial recognition services to assist despots destroying the lives of humans.

How much money is enough Bezos? Does your toy rocket make you feel better about it? He probably wasn't so bad at one time. It can happen to any of us even without masses of money to lubricate the process.

Nonbinding resolution saying "Don't sell this stuff to evil bastards." Non-binding. Can't even have that as an "ideal" to aspire to. Get rid of that, could be money in it.

[1] https://en.wikipedia.org/wiki/IBM_and_the_Holocaust

I'm pretty sure Microsoft sells facial recognition to government and military. They have opposed some bills which imposed too strict restrictions on its use as well if I reckon.

And Amazon has similarly suggested that there should be regulation put in place.

Here's the thing, regulation is good in this case, as long as it doesn't prevent Microsoft or Amazon from still selling the tech. That's why Microsoft opposed certain proposed bans, but recommend for regulation.

Things like, having to get a warrant are good, because Microsoft still sells the tech, but they don't get blamed when its use wrongly.

Do you have a source? I’m aware of Microsoft military contracts ie HoloLens, but have no knowledge of facial recognition contracts.

Also source for amazon suggesting regulation?

For Amazon, I think that's what I'm thinking of: https://www.wired.com/story/amazon-joins-microsofts-call-rul...

For Microsoft, I can't seem to find what I had read anymore, all Google result seem overshadowed by articles of them "calling for regulation".

I did find these:

https://www.wired.com/story/microsoft-wants-rules-facial-rec...

https://www.businessinsider.com/microsoft-cruel-to-stop-gove...

https://www.businessinsider.com.au/microsoft-china-muslim-cr...

Interesting. Thanks for the context.

What was the point of taking this to shareholders? On what planet would anyone expect them to vote any other way?

The point of it was to deliver an excuse as to why their hands are tied, it went to vote precisely because the outcome was already obvious.

Almost all blue collar employees (software engineers) are shareholders at Amazon.

> blue collar employees (software engineers)

Huh?

Working class might've been a better term

SE are not working class. At best they are at the top end of the middle range, and generally above 100k is considered 'upper middle'.

The idea of SE as working class is from categorization based on labor relations, not on income.

In a Marxist-theoretic framework, for example, SE are clearly working class.

Software engineers quite often have and derive substantial income from capital holdings; when they do they are, in a Marxist framework, of the petit bourgeoisie (the middle clsss between working class [proletariat] and major capitalists [haut bourgeiosie].)

But, yes, some software engineers are working class, as well, the profession largely being split between the proletarian intelligentsia and petit bourgeois intelligentsia.

Software engineers are white collar workers[0]. As far as I know, hourly employees no longer get equity[1], and I would guess they comprise most of Amazon's blue collar workforce.

0: https://en.wikipedia.org/wiki/White-collar_worker

1: https://www.cnbc.com/2018/10/03/amazon-hourly-workers-lose-m...

Sorry, working class (as some have said) or white-collar would have been a better choice of words.

That is precisely the point: to externalize and diffuse responsibility on an abstract concept ("the market") rather than hold its creators accountable.

A tiny fraction of shares are probably owned by activist individuals with socially responsible interests.

The rest are employees [1] with self-interest and giant funds with big piles of them whose only metric is returns. Of course they'll all vote for the gain.

https://www.investopedia.com/articles/insights/052816/top-4-...

~36% of the voting power comes from top 20 institutional/mutual fund holders and about 14% of the vote comes from Jeff Bezos himself[1]. The board of directors recommended all holders vote against this. Even if every shareholding employee of Amazon voted for this, there's very little chance it would've made it past the institutions voting along with the board recommendations.

[1]: https://finance.yahoo.com/quote/AMZN/holders?p=AMZN

It amazes me still how much power in our "democratic" system is help completely outside of the control of the public. How some people, by virtue of owning capital, are vested with such powers that can make or break the lives of millions. Theirs are the decisions that can kill thousands or cause suffering to millions.

Democracy is not a boolean. It's a continuum. To me, it means "power held by the people whom it concerns". That's it! But even in our best democracies, only some portion of power is held democratically. Other decisions, often no less impactful or critical or harmful (think polluting factories, advertisers that insidiously manipulate people's psyches, governance of enterprises bigger than many nations, etc), are done by undemocratic, unelectable, unaccountable elites.

This needs fixing.

PS: this is bad even if the capital "well earned", for some definition of "well earned", but it is especially aggravating when we're talking about the Saudi royals or the Waltons of this world...

Uncanny. I watched the documentary "The Corporation" the other day (https://www.rottentomatoes.com/m/corporation) and it opened my mind to this issue completely.

It is well worth the watch.

Huh? If the government wanted to ban facial recognition, they could do it. Our democratic system absolutely has the power to ban it. It hasn't been done, presumably because there isn't a public consensus that it should be banned.

If you're arguing for public ownership of companies, that's literally communism -- so not "democratic." But we don't need communism to regulate facial recognition.

If you hadn't replied first, I would have replied the exact same thing as your first paragraph - also starting with "Huh?".

That said, public ownership of companies is not necessarily undemocratic. It's just a bad idea for different reasons. USPS would be an example in the US of publicly owned company.

I believe that Marx even originally intended communism to be democratic. The regime in the USSR was supposed to be transitory.

Exactly. Does Amazon have an army more powerful than the US? No? Then absolutely the government can regulate corporate America. It's a matter of the will of the people.

Edit: 20 bucks OP (andrepd) has an Amazon Prime account.

>Absolutely the government can regulate corporate America. It's a matter of the will of the people.

The issue (or rather, one issue among many) is that that isn't so straightforward. There is an inherent assymetry: a trillion dollar corporation can influence the government at a scale so staggeringly huge and pervasive there isn't even a comparison that can be made with the power and influence of a regular Joe. They can lobby government, bribe politicians, run ads, put on massive campaigns, all of which distort government into their hands.

>Edit: 20 bucks on OP (andrepd) having an Amazon Prime account.

Cute of you, but no. I've bought things from amazon 4 times since 2015 (I just checked).

That is why cable modems hold special status via FCC rules and thus federal law (in the US).

>There is an inherent assymetry: a trillion dollar corporation can influence the government at a scale so staggeringly huge and pervasive there isn't even a comparison that can be made with the power and influence of a regular Joe.

Where does Amazon's power originate from?

A Princeton study in 2014 found that public opinion in the U.S. has virtually no influence on national policy decisions.

https://mavenroundtable.io/theintellectualist/news/study-the...

Subsequent studies have found that it's not as bad as all that. But even they found that the public often doesn't get what it wants. Therefore we can't conclude there's no public consensus for a ban just because the government hasn't done it already.

A Brief History of Neoliberalism, 2005:

http://rebels-library.org/files/d3Thompson-1.pdf

Well, yes and no.

Actually, for US public companies, about 60% of value (common stock) is owned by individual citizens, including beneficial ownership through mutual funds. Everyone with a bit of retirement savings. And that proportion is actually increasing, as institutional money chases private asset classes.

The real democratic tragedy is that these retail investors never vote (less than 11% of accounts vote at all). So results like this one are dominated by hedgefunds, who are the most vocal voice in the room. The big mutual fund managers often abstain, in large part because they don't know how their constituents want them to vote.

So there is actually a big democratic opportunity -- the public owns the majority public companies, and has the ability to hold them accountable to public interests. If only people would pay attention. Usually, people throw away their proxy voting cards.

To try and address this problem, I built this website, built around a simple online petition UI (a la change.org) but which verifies your shareownership (via api to service like Plaid). For example, one of our users put up this exact petition on our website. https://www.yourstake.org/ask/amazoncom-inc-facial-recogniti... We've been hacking away on features recently, and haven't yet gotten the word out enough.

But as a 401k or Roth mutual fund, the citizen could be in our out of the particular stock at any given moment, without their say.

I get what you are saying, but this gets really tricky really fast.

The power you are talking about is the power to use what you own (in this example, that thing is money). If we believe in private ownership of things, we will never be able to eliminate this power completely. Where do we draw the lines? How do we let people have the freedom to choose how to use the things they own without letting them have too much power over others?

I think we can come up with good answers, but I do fear an over simplification of what it means to restrict how you can exercise your ownership of something.

Sure, Amazon shareholders can "make or break the lives of millions"... but only if the government lets them. Neither Amazon nor its shareholders can really overrule what the American government decides. In theory, this means that you and I hold ultimate power, and Amazon operates at our discretion.

The real problem is that the American government is beholden to corporations like Amazon, and the ultra wealthy like Amazon's shareholders. Not you and me. This in turn comes back to campaign finance reform. As long as politicians are beholden to the wealthy to fund their election campaigns, you and I aren't going to have nearly as much say.

What I really wanted was facial recognition in Google Glasses. I hate it when I am in the store and someone comes up and says Hi <name> hows the wife and kids, and I am like "who is this....". Would be great to say "Hey Bob, doing good" :)

I'm sitting here imagining a very distant future where the part of the brain that recognizes faces eventually becomes vestigial and we end up relying on technology to recognize the people in our lives.

Basically acquired prosopagnosia.

Almost feel like that's not so distant. Imagine how fewer people we would recognize if we weren't regularly pinged with an update about them or their picture scrolled passed on a device.

That sounds like the tech helping memorize faces, the exact opposite?

Soon you will also be able to see a rating next to their name showing their social credit score so you will know how much interacting with them will affect your score.

Hello Black Mirror.

And that is good news. Company should focus on making good services and good profit, not virtue signaling on political fluff

It is not "virtue signaling" to choose to stop doing evil.

Just a reminder that like encryption, facial recognition is just applied math. Why are people trying to ban math?

First off, this was specifically about banning the selling of facial recognition software to specific players.

Also: Just a reminder that nuclear bombs are just physics is just applied math (kinda).

Just because what you're doing is rooted in the field of mathmatics doesn't automatically mean you should be doing it or selling it, especially when your application is crossing into the physical world where there is the wellbeing of humans to be concerned about.

Even if the entity you're selling to could just very well build it themselves, it still doesn't remove your moral obligation to not be the one providing it.

Regulating this technology seems like regulating encryption. Futile. You can never put the tech genie back in the bottle. Especially since presumably many of the foundation layers of this technology are open source with wide scale applications. Will they outlaw machine learning? TensorFlow?

Obviously whatever Amazon is doing is more complicated than throwing a ton of face images at TensorFlow but as technology progresses, the advances will bubble up to the application layer.

The thought that you could even regulate facial recognition just shows how out to lunch lawmakers and the media are when it comes to technology, how it's developed and how it's deployed.

This isn't about regulation of facial recognition tech (in the legal sense), this is about voluntarily choosing to not pursue a course of action because it is unethical and how shareholders have overruled decision makers on that matter.

From the article:

Responding to the shareholder vote, Democratic U.S. Representative Jimmy Gomez said, “that just means that it’s more important that Congress acts.”

Republican members of the committee expressed concern about U.S. citizens having their rights violated and their information shared with the likes of the FBI without elected officials’ oversight. A second hearing on the topic is scheduled for June 4.

Any federal regulation would follow a major decision by San Francisco officials earlier this month to ban city personnel from using the technology. That contrasts with New York, Chicago and Detroit, where law enforcement have reportedly used facial recognition or acquired it with the hope of speeding up post-crime investigations.

Actively producing and selling the tools to strip human rights is far worse than standing by while someone uses said tools.

Not all uses of facial recognition strip people of human rights. Would it strip you of your rights if an airport used this tech to discover known terrorists in a crowd? Or how about recognizing you at the DMV and pulling up your records automatically? I'm sure there are millions of non-human rights violating usages (even by the government) of facial recognition.

IMO it isn't that it strips people of human rights per se, it is that it removes opportunity to disrupt the status quo. Whether you believe this is a good thing is up to you. Terrorists are trying to disrupt the status quo, but so are protestors, people who write to their Senator, etc. By the ubiquity of facial recognition, you are essentially making it harder to disrupt the status quo by non-legal means. This all has the effect of increasing cohesiveness of the existing system. When people talk of "dystopia", that doesn't necessarily mean it is a bad place to live, at least in principal. Brave New World was tyrannical and oppressive, but not necessarily a miserable existence (you always have SOMA). What distinguishes this as a "dystopia" is the inability for the system to change itself. So while it doesn't strip human rights, per se, it certainly strips people of the ability to affect change on their environment. If knowledge is power, then the more knowledge they have about you, the more power they have over you.

Technology empowerment cuts both ways though. If you regulate technology so only the government has it, then the scenario you lay out will still happen. If you leave access to technology available to everyone, those same protesters could deploy their own version to locate government moles in their midst (for instance).

Technology has the ability to power everyone, individuals, groups and governments. It's best to leave the access open to all.

> Obviously whatever Amazon is doing is more complicated than throwing a ton of face images at TensorFlow but as technology progresses, the advances will bubble up to the application layer.

You’d be surprised...

> Regulating this technology seems like regulating encryption. Futile. You can never put the tech genie back in the bottle.

We've (effectively) put chemical warfare back in the bottle, regulations can be quite effective in shifting the public perception - but I won't argue that it's hard to do.

Given tear gas I am not sure about chemical warfare (although that also has to do with relative effectiveness).

And there is far more infastructure required for chemical weaponry production and use more complex than throwing a bottle of bleach and ammonia mix.

Chemical weapons have no consumer applications and are very difficult to create. This is more like trying to put DeCSS back in the bottle:

https://en.wikipedia.org/wiki/DeCSS

"The proposals faced an uphill battle. Amazon’s board recommended against them" - for those not aware, the way shareholder voting works for Amazon, is that shares that aren't voted are assumed to vote the same way the board recommends. This means to get any proposal different to the board's recommendation, you need to get more active voters for the proposal, than those opposed + not voting.

There was a shareholder proposal recently to make the default for non-voters abstain instead of "whatever the board wants", but the board recommended against it and it was defeated.

What’s the backstory here? Did Amazon ban some facial recognition product it’s developed from being used in certain circumstances and the shareholders overturned that decision?

No, some shareholders brought forth a motion to ban the product at the shareholder's meeting, and it was voted down.

Security cameras are already everywhere. So, um, should there be a law that only human retinas can look at them? Why allow recording of the feeds, for that matter?

Why would shareholders reject something that can make them money? This is not a charity.

Conscience?

What is wrong with facial recognition ? The problem should be government using it to enforce laws. Even that is debatable. Most public spaces are currently under surveillance.

Dont buy from Amazon.

How hard can it be

Surprisingly difficult. Winding down from Amazon personally has taken months, but thats futile. At work I spend 8 figures a year on AWS.

And there are no alternatives that exist?

AWS definitely has competitors you can switch to pretty easily, I think Amazon's retail side is the harder portion to jettison.

Do you use AWS? It's extremely difficult to move away if you use services like RDS, ElasticSearch, ElastiCache, Lambda, KMS, parameter store, etc. It will cost a couple years of ops time to reliably migrate to a competitor like GCP + change work patterns. Not to mention that Google products like Functions are still alpha and unreliable.

The retail side is very easy to jettison. Most of the goods can be found at target, Walmart, Best Buy, Costco, or Newegg. At least it was for me.

For example, my friend has a doorbell with a camera. The doorbell has software that runs facial recognition on me, and it often correctly identifies me, regardless of whether I look into the camera or not. (Nowadays I do, since I know it already knows me, I just make silly faces since I know it goes to his phone.)

Now, IDK how I feel about this doorbell. But it's also his property; am I to say he can't have an AI-enhanced doorbell on his house? But that doorbell also likely uploads my image, and data about me, to centralized servers somewhere, largely without my consent. (Aside from perhaps my implied consent of entering his property, but even then, I don't think I noticed it the first time I visited. Now I'm aware of it, since he told me, and has demoed the tech to me in various ways.)

(IDK if this particular thing is an Amazon product or not, and in fact, I highly suspect it isn't, but I want to pose it as an example of just because I might voluntarily choose to not use the tech, that choice only goes so far, and there are still privacy implications for me despite it.)

> Now, IDK how I feel about this doorbell. But it's also his property; am I to say he can't have an AI-enhanced doorbell on his house? But that doorbell also likely uploads my image, and data about me, to centralized servers somewhere, largely without my consent. (Aside from perhaps my implied consent of entering his property, but even then, I don't think I noticed it the first time I visited. Now I'm aware of it, since he told me, and has demoed the tech to me in various ways.)

Even if it weren't his property, if you were in the public aren't people legally allowed to take photos of you and upload it wherever they want? If it weren't legal there wouldn't be paparazzi I would think.

> if you were in the public aren't people legally allowed to take photos of you and upload it wherever they want?

No, not in the US, depending on exactly what you're doing. Taking pictures of people in public is legal, but there are restrictions on what you can do with them.

> If it weren't legal there wouldn't be paparazzi I would think.

The law is different, and less restrictive, when it comes to certain classes of people such as celebrities and politicians.

> Even if it weren't his property, if you were in the public aren't people legally allowed to take photos of you and upload it wherever they want? If it weren't legal there wouldn't be paparazzi I would think.

This is probably one of the areas where the law's not quite caught up to reality.

50 years ago, the probable life story of a picture of you taken by someone in public was that it would be developed, maybe reproduced once or twice, possibly sold to a magazine or displayed somewhere, and then it would spend the next half century in a drawer somewhere.

In 2019, the life story of a picture of you taken by someone with an Android phone is that it is uploaded to Google's servers, where it's analyzed, tagged via facial recognition, combined with location data, associated with your profile, and used to build a better understanding of who you are, who you know, and who you associate with, so Google can package that data and use it to target better ads at you (and Google's one of the "better" players in this space, in that the worst they're likely to do is advertise to you).

I don't think it's quite caught up to people that the average amount of processing being applied to any given picture of them in 2019 is about on par with what the CIA would've dedicated to a picture of the Russian ambassador back in 1969.

How does the GDPR apply in this case? If they take a photo of you in public and manage to associate it with your profile, I assume that instantly becomes PII that you have a right to have deleted, right?

First you have to even know that your data was taken and which company it was uploaded to.

Actually pretty ease with Ali-Express, but I don't know how much that fulfills your reason to not buy from Amazon. The brick and mortar stores are probably never coming back like they were before.

Not hard.

But boycotting Amazon because of it is selling facial recognition API? Most people will call it nonsense.

Looks like Amazon is incapable of policing itself then.

Progressives are progressive until it's their own backyard. Great article about that earlier today on HN.

Do you have a link handy?

I don't understand this comment. Being critical of the implications of facial recognition technology doesn't map cleanly along the political spectrum, but the most vocal criticism has come from the left. Amazon shareholders largely consist of institutional investors whose interests (and politics) tend to the more conservative.

In what sense are progressives being hypocritical here? As far as I can tell progressive (leaning) employees and other minority shareholders proposed an initiative that was rejected by the majority, more conservative shareholders.

Ah, the public corporation! Fictive structures created so that profit-making shall never be interfered with by moral or ethical concerns!

Just creating shareholder value, what else can you expect us to do!

I do wish you were exaggerating...

Queue the torrent of HN comments that conflate "is technically legal" and "Is morally responsible", after all the highest form of morality is making money.

It's always funny to see the upvote/downvote swings on comments like these.

I try to keep them more substantive overall, but what else can be said about this. Clearly there are still major ethical concerns with access to these facial recognition black boxes and clearly amazon shareholders are far more concerned with expanding into any legal market as quick as possible.

Amazon is /the/ poster child of late capitalism as far as I'm concerned. Ruthlessly efficient, impossible to escape even if you hate them (I mean, come on, like I'm gonna go to a store when I just grab it on Prime now), and literally run all the retail stuff as a revenue front to expand deep into subterranean territory (AWS, etc). No nod to "Don't Be Evil" here, just a gigantic black box that can pump out money for shareholders. Enough money and leverage to apply impossible pressure in almost any market they decide to want to dominate (Still waiting to see what they do to grocery with the WF buy).

And still mostly building the layers and connections for domination. They haven't even switched into eat-the-world mode yet, they are still just grinding and building resources...

> "Amazon is [...] ruthlessly efficient..."

this part is capitalism working as intended and (mostly) as desired. the challenge of course is that it has also led to highly undesirable power structures (corporations, oligopolies, cabals, cartels, etc.) that increasingly eclipses the individual as principal agent. that's a failure (by omission) of capitalism.

while governments, communities and other social structures can contain that consolidation of power to some extent, subservience to corporate power is a looming threat for common folk. it's crucial to further codify and protect the equivalence of the individual and the corporation.

for instance, contracts (like EULA's) shouldn't be one-sided ultimatums inflicted on the indivudual, but the beginning of a negotiation between two equal parties. corporations would argue that the administrative overhead of this would be prohibitive, but that's the price for access to a lucrative market (and that sweet, sweet revenue stream).

in addition, corporations naturally have asymmetrical information advantages that need to be leveled in the exchange of value contemplated by an economic transaction. we have financial disclosure laws, for example, but they don't go far enough. basic information like the actual prices paid by other customers and the reasons for differences should be made available to the counter-party.

>> "Amazon is [...] ruthlessly efficient..."

>this part is capitalism working as intended and (mostly) as desired.

I think we need to be careful to define what we mean by "efficient". Do we frame efficient as "churns out the highest revenue / spending ratio" (ironically evaluating capitalism by a metric created by capitalism)? Then yes, that may well be true. But what about the overworked and abused workers? What about the stifled competition? What about the 40% increase in fuel and emissions to reduce shipping from 3 to 2 days? Is it all worth it, to get your consumerist crap a few hours earlier in the post? This are not just rhetorical questions, these are actual questions we need to pose ourselves (e.g. is it worth an excess X million tons of CO2 per year to get prime deliveries Y hours faster).

Would that fit into a definition like "increases overall quality of life the most"? Food for thought, in my opinion.

>for instance, contracts (like EULA's) shouldn't be one-sided ultimatums inflicted on the indivudual, but the beginning of a negotiation between two equal parties.

Isn't this one of the classical critiques of capitalist (e.g. Marx), that it is impossible to have a negotiation between two equal parties when the two parties are so... well, unequal.

Maximum efficiency will occur when no humans are employed and all energy is produced and transmogrified for, and consumed by, retrofeudal neoaristocrats living lives of infinite leisure in their garden cloud paradise fortresses.

Free markets are neither entirely good nor entirely bad. Many things that we would consider to be intrinsic and inextricable parts of the human experience are things which cannot survive the relentless assault of unrestricted Darwinian competition.

A free market is a defect-defect equilibrium, and a defect-defect equilibrium is like fire. It's very powerful, and absolutely necessary; we wouldn't have what we have without fire, and we wouldn't even be human as we are without it, but we must be very careful that it doesn't escape the fireplace and burn the house down.

You ask what would "increase overall quality of life the most?" There are non-economic markets, as well. Mating, for instance. Until very recently it was highly monogamous and strictly regulated. And now it isn't. Winner-take-all.

I don't know where I'm going with this, except to say that our world is the direct product of decisions, our collective decisions... and the people who make them:

"The conscious and intelligent manipulation of the organized habits and opinions of the masses is an important element in democratic society. Those who manipulate this unseen mechanism of society constitute an invisible government which is the true ruling power of our country. ...We are governed, our minds are molded, our tastes formed, our ideas suggested, largely by men we have never heard of. This is a logical result of the way in which our democratic society is organized. Vast numbers of human beings must cooperate in this manner if they are to live together as a smoothly functioning society. ...In almost every act of our daily lives, whether in the sphere of politics or business, in our social conduct or our ethical thinking, we are dominated by the relatively small number of persons...who understand the mental processes and social patterns of the masses. It is they who pull the wires which control the public mind."

— Edward Bernays, Propaganda.

> 'I think we need to be careful to define what we mean by "efficient".'

yes, efficiency has multiple definitions (e.g., productive, allocative, etc.), but i think the problems you raise (overwork, anti-competitiveness) are primarily regulatory. communities (governments) set the rules, and capital has to be allocated within the confines of thsoe rules, like 40-hour work weeks and antitrust laws (akin to out-of-bounds rules in sports). similarly, the costs of pollution should be internalized, as logically, it's a form of stealing (taking enjoyment, health and even lives from others).

> "Isn't this one of the classical critiques of capitalis[m]..."

i suppose marx made a similar argument about the tension between capital and labor, which could be further extended to the consumer-citizen as another economic stakeholder. we invented the fictional personhood of the corporation to resolve a legal conundrum (how to interact legally with businesses, separate from owners) and even gave them special privileges to foster innovation (e.g., the corporate veil). it's only fair to maintain equivalency in negotiations in exchange for those privileges.

Should also add that maximizing shareholder value isn't a hard legal requirement since those laws are often misquoted to imply the opposite for some reason and that the board had "no choice" to pursue a particular controversial revenue stream, source: https://www.lawschool.cornell.edu/academics/clarke_business_...

Yes. If maximizing shareholder value was the end-all and be-all, then every corporation would be in the sex, drugs, and gambling businesses.

I mean, the gambling business may be profitable, but are the profits really left on the table at this point? You don't just have to build a casino, you have to build a /better casino/.

Profitability has little to do with the overall margins of the market you are in, but in whether you can find places to optimize ROI that your competitors have not.

Gaming industry has around 10% net profit margins. Compare that to Apples 30%.

Why do you deem all use of facial recognition to be immoral? Is it more moral to only allow the governments access to this tech? Are there really no consumer friendly applications of facial recognition?

I'm sure there are and the only way we can explore these benefits is to allow the population access to the technology.

RTFA:

> The first proposal would have made the company stop offering facial recognition to governments unless its board determined sales did not violate civil liberties. A second would have requested a study by September of the extent to which Amazon’s service harmed rights and privacy.

There is no world in which you are not disgustingly greedy, unempathetic, and immoral to reject sales of facial recognition technology to abusive governments.

Whether a governement is abusive or not is greatly in the eye of the beholder so yes, there are plenty of worldviews where such a thing wouldn't be "disgustingly greedy, unempathetic, and immoral".

Besides, there's a lot more going on in this space (and article) than the vote itself. There are calls to regulate facial recognition, which I think is an immoral option (and futile) as it only allows governments access to a powerful technology.

I think it’s entirely misguided to place the responsibility to ensure that customers are not using products to violate civil liberties on to corporate boards. Legislature should define them, and courts should uphold them. To say that these responsibilities should be shifted to the board room is alarming and undemocratic.

That's not what the rejection of this proposition indicates. The shareholders are not saying, "Amazon, sell to anyone no matter how bad." They are saying, "Amazon, continue using whatever processes you are already using to determine who to restrict from using this product."

It is very common for people who have a particular issue that they care about to attempt to require the board of a company to handle that issue. But if big companies acquiesced to all such requests, the board would take on a large workload, making sometimes detailed decisions where they have no expertise.

Company owners delegate much of their authority to boards. Boards delegate much of their authority to employees. I am sure there will be entities to which Amazon would not offer facial recognition services. But that decision does not need to be made at the level of the board. BTW, this is why management almost always fights such propositions: it ties the hands of the company to use an inefficient decision process, for no provable gain on any worthwhile metric.

you're argument is wholly misplaced.

if it were a decision like "should we put this bin over on that shef?", then yes, it's reasonable to delegate to a rank-and-file employee.

"should we develop technology that can lead to a surveillance state?" is not one of those decisions.

that's a strategic decision entirely negotiated between the shareholders (via the board) and the management team, for which they should all be held accountable.

Almost all useful technology could conceivably lead to better surveillance. Email is a great example. It is infeasibly expensive to open and read everyone's physical letters. But reading everyone's email is possible and is literally done in some very large countries.

I believe that if the board of Amazon had more they wanted to say about this issue, either publicly or internally, they do not need a shareholder resolution to enable that -- especially a non-binding one. They can simply do so. So, the resolution's purpose is actually one of signalling the shareholder's stance on the technology being developed. Clearly, most shareholders agree with the board's implicit approval.

> I am sure there will be entities to which Amazon would not offer facial recognition services.

I see no reason to suspect that this is true.

>There is no world in which you are not disgustingly greedy, unempathetic, and immoral to reject sales of facial recognition technology to abusive governments.

You need to re-word this sentence. I agree with what (I think) you're going for, but your words say something else.

There are many people who don't think there's anything immoral about facial recognition, or even its lawful use by law enforcement. I am among them. Such individuals probably comprise a higher fraction of Amazon's voting shareholders than the population at large. Especially if you count by votes, since if Bezos thought there was anything wrong with it the program wouldn't exist.

Just out of curiosity -

How have your interactions been with law enforcement during your life? Largely positive? Largely negative? Or non-existent? Do you, overall, feel that the US in largely a place where law-abiding citizens are free to go about their days?

There are obviously serious problems with law enforcement in the united states. Although I think that these mostly manifest in face to face encounters with the poor, people of color, and the mentally ill. I can't think of many recent cases where local law enforcement's application of information technology resulted in gross injustice.

The idealistic argument would be that moving to IT and algorithms reduces the gross amount of human bias that currently enters into the system.

The reality may be far different as the algorithms themselved make the biases even less seen and less able to be contested.We're not in full minority report territory, but I think the worry is that added the ability to do near permanent surveillance to an already militarized police force that thinks they are fighting a war is not...going to turn out well for some of us.

> The idealistic argument would be that moving to IT and algorithms reduces the gross amount of human bias that currently enters into the system.

That is not my argument. I am merely saying that I don't think facial recognition is going to present problems that outweigh its value as a law enforcement tool. There will no doubt be some abuses, just as there is of any new technology. I support regulation and intense scrutiny of how law enforcement uses these tools, to minimize the severity and duration of these abuses. But I don't think that means it's a good idea that Amazon's board should have to approve each entity attempting to use their services individually.

I’ll respond too, since I agree with the guy you’re responding to.

By and large, I distrust American cops. I’ve had some bad experiences with them (though never catastrophic) but I’m well-aware that some people have had it far worse.

I don’t have a problem with facial recognition software being used by police. I have concerns about equipping police with MRAP vehicles, flashbangs, tasers, body armor, rifles, tear gas, and a few dozen other things, but not facial recognition. Why should I be concerned?

So it is ok that China is using Facial Recognition and active surveillance to oppress, as long as it is not in your backyard?

https://www.nytimes.com/2019/05/22/world/asia/china-surveill...

I’m not in favor of doing business with the Chinese government. Pretty sure the Chinese are capable of developing the technology themselves, anyway.

All positive. Any negative experiences were in court, not the frontline police patrols. There are many layers to enforcement and it might be surprising where exactly the biggest problems are.

Adding increasing automation when the foundations aren't great is a bad idea though. There are more stories everyday about people finding trouble with the law because of some software system with no human oversight.

Public corporations, private corporations, state-owned, party-controlled corporations. It doesn’t matter. At least public corporations are known to bow to pressure.

I think ownership is irrelevant. What matters is lawfulness and what remains lawful after pressure or indifference from the constituencies.

Hand in hand with this is the rule of law. Some countries on paper have great liberties and rights for their people, but in practice it’s up to the party to interpret and enforce (as in the great purges, etc,)

I would agree with you entirely and I think that the history of both capitalism and communism have demonstrated that we are not at a level yet where we can design a socio-economic system to make it not matter whether we are good to each other and in fact, they are both likely, in different ways, to sometimes distort and amplify some of the worst parts of our animal natures.

I have found it a certainty on HN that at some point, in any discussion of these things, that someone will step in and say "but this is legal, and the corporation must maximize its profits for the shareholders, so if you really want a change, change the laws, the company itself is utterly amoral and working as designed"

> What matters is lawfulness and what remains lawful

To expand on that idea - if something is both lawful and profitable, you have to have an almost absolute consensus amongst the people capable of doing it that it should not be done. Otherwise, potentially even if 1 person thinks it is fine, then they will be the one to do it.

Typically this is the main strength of the economic system - jobs that people don't want to do get allocated to the person who has the least problem with doing it.

The threat here was never the development of the technology - it is obvious that it is coming - the issue here is government demand. A corporation can only do so much with my face that annoys me - the government on the other hand can make life really miserable.

> if something is both lawful and profitable, you have to have an almost absolute consensus amongst the people capable of doing it that it should not be done.

Sure, but you just need a weak consensus amongst everyone that it should not be done to make it unlawful.

There are B-Corps (benefit corp) in many states that have charters that include bylaws that are as important or at least equally as important as profitability. IMO all health providers should have to be B-Corps that have patient health as their #1 above profit.

"B-Corp" is a certification[1]. It has no force of law anywhere. It has nothing to do with a benefit corporation[2], which does.

[1] https://en.m.wikipedia.org/wiki/B_Corporation_(certification...

[2] https://en.m.wikipedia.org/wiki/Benefit_corporation

It's always good to remind: lawful != moral/ethical. Sometimes the law is twisted and wrong, and it needs to change.

Very true. And ethics change over time. Laws change over time. But people want immediate reconciliation with their beliefs —which can be at odds with that of others. So we have to rely on the law.

Prohibition is a great example of what laws can do, good and bad. After lots of bloodletting we realized it wasn’t good and we were not benefitting from it so it was reversed.

Governments are fictive structures created, in part, so that they can enforce moral and ethical concerns against profit-making. That's where attention should be focused instead.

Shouldn't this technology be outlawed outside law enforcement?

It sounds like it can be used like a powerful tool.

Why should private analysis of photos legally acquired be illegal?

If computers are legal and software is legal and photography is legal, it seems to me that running your own software on your own computers over your own photos should be legal.

I don’t like the implication either, but I think perhaps the thing we should be looking at is why it is legal for companies to take photos of us for commercial purposes without our explicit consent. I also feel it would be much more practical to regulate and enforce than regulating what algorithms they get to use on already-captured imagery.

Right, so parent is suggesting making facial recognition software (as a class) illegal. You'd still be able to use your legal computer to analyze legal photographs, but we could, as a society, say it is a Bad Idea™ to make that analysis trivial.

Alternatively we could say that everyone should be able to do facial recognition analysis, but then access to the photographs becomes the gatekeeper of power.

Amazon stuff runs on Amazon servers... cloud means, it is not your computer, most likey not your Software and sooner or later not your data.

> Why should private analysis of photos legally acquired be illegal?

For the same reason most horrifyingly destructive items created out of perfectly legal items are illegal?

Is the sole argument that you're presenting in favour of outlawing really that 'it can be used like a powerful tool'?

I'm actually rather the /most/ concerned about law enforcement having access to it.

Yeah banning its use outside law enforcement sounds like the worst option given it is anti-transparency and accountability while allowing abuse. It woukd be far easier to selectively hide misconduct that way.

Yeah I don't care much about what Amazon does with it, at least compared to the police. Amazon doesn't have the ability to kill me or lock me in a cage for 50 years.

A really good definition of government that I've heard is "an organization that has a monopoly on force." There's probably more that I'm forgetting but I think that definition is important because it's something that people often forget or choose to ignore.

> A really good definition of government that I've heard is "an organization that has a monopoly on force." There's probably more that I'm forgetting but I think that definition is important because it's something that people often forget or choose to ignore.

Max Weber, although his definition is a bit more nuanced:

https://en.wikipedia.org/wiki/Max_Weber

> Weber defined the state as an entity that successfully claims a "monopoly of the legitimate use of physical force within a given territory".

"Legitimate" is a tarpit, but in practical terms it's pretty much what others recognize as legitimate. That's the difference between a warlord and a dictator: One controls an armed gang, the other controls a military.

This poses some problems for people claiming that anarchism can work: If a region declares itself anarchist, but the majority can and will enforce its rules by force, is it anarchist or majoritarian? Is there any practical difference?

> Amazon doesn't have the ability to kill me or lock me in a cage for 50 years.

I wouldn't be so certain about that. History is filled with examples of corporations doing exactly that sort of thing, usually (but not always) by having a government act as their proxy.

Reply on news.ycombinator.com