Google Workers Urge C.E.O. To Pull Out of Pentagon A.I. Project

Sometimes the distinction is even more insidious. I did work on perpetual flight for drones, and Facebook had a perpetual flight project that had the goal of bringing internet access to remote locations in Africa. Sounds humanitarian, but I also didn't want to be responsible for subjecting poor Africans to what I consider Facebook's panopticon. Or maybe it would have been a net boon for the region? It's really hard to tell a priori, so the best I have managed for myself is just to try and stay in theory, where developments are further removed from direct consequence.

In the case of Fritz Haber, it's the same man who saved billions from hunger and killed millions by prolonging World War-I. Excellent article on the same - https://medium.com/the-mission/the-tragedy-of-fritz-haber-th...

Sometimes, the decision isn't clear as black & white; especially when you mix nationalism & patriotism into it.

So, there are two aspects of that: first, do you trust the people you're designing a weapon for to use it for justifiable purposes, and second, how does the weapon itself influence whether it can be used for just or unjust reasons? Gas chambers, for example, can only be used to kill someone whom you are already able to coerce into entering it; its only use is to murder a helpless prisoner. A Spitfire or Hurricane can shoot down bombers but not necessarily do that much damage on the ground. Orwell had an interesting perspective on this question: "Thus, for example, tanks, battleships and bombing planes are inherently tyrannical weapons, while rifles, muskets, long-bows, and hand-grenades are inherently democratic weapons. A complex weapon makes the strong stronger, while a simple weapon — so long as there is no answer to it — gives claws to the weak."

I really don't have answers. The stance of, "weapons equals bad, therefore it's wrong to ever help design weapons" is a pacifist notion, and maybe that appeals to you, but if you're not a pacifist, it becomes a complicated question with certain political ramifications.

If one could engineer a solution to save lives, but instead uses his/her talents to build the nth iteration of, say, a food delivery app, does he/she bear any ethical culpability?

Because it seems the bigger challenge for engineers is not avoiding overt evil, but in the opportunity cost of spending their time on work of significantly lesser consequence than their talents might allow.

http://ccr.sigcomm.org/archive/1995/jan95/ccr-9501-clark.pdf Here they say that the design of the internet was very much determined by the military priorities of the project (Telcos had different priorities so they came up with OSI where the link layer is supposed to be reliable, as opposed to DARPANET/IP, where the link layer/IP routing is dumb and the endpoint is smart; this allowed link layer to be pluggable and IP routers to be cheap)

I used to find myself most of the times confused while using technologies whose CEOs, mission I do not agree with. It felt as if I was contradicting myself. But then I realized the dialectical nature of this relationship which is if you want to create something better, improve something it is only natural that you will do so using the existing tools available in your world.

A great example is Facebook. There are things you might not agree with, you might hate it but if you look closer you will see volunteer groups to political campaigns to marches and protests which are organized only via Facebook. One of such political activities one day could as well change Facebook or the world we live in drastically.

You can apply this to other cases. You might have an anti-capitalist philosophy podcast hosted on Amazon or Spotify or Apple which in essence contradicts the very existence of such companies. But in the end you are using them to deliver your message, cause change etc.

I am curious about what others think about this and how they feel about looking at it from a dialectical point of view as I am continuously thinking about this and still forming my thoughts.

But there is space for ethical contemplation during an arms race.

That's the truth.

Did the engineers of the atomic bomb save millions of lives and usher in an era of unprecedented peace? Or did they slaughter millions of innocents and sow the seeds of humanity's destruction?

It's been 70 years; we've had plenty of time to make up our minds.

Many inventions are not so easily classified as innately good or evil.

>Kearns was a member of the Office of Strategic Services, the forerunner of the U.S. CIA, during World War II.[1][2][3]

Somehow I don't think that Kearns would have had an issue with wriing software to analyze video footage for the military.

Merely offering a force projecting capability is not evil.

If you look at world history, the only concrete definition of who has most political power is this: it's the party with the most capability for projecting violence in a given area.

Small close knit groups that live outside of large populations can live independently and do without military force. Once you have a large population with more or less fixed hearth and home the party with most violence projection capability owns them politically.

This dynamic is played in almost ridiculously minuscule scale in the peloponnesian wars of the greek city states of antiquity.

The pathological manifestation of this principle can be viewed, say, in the rise of ISIS and the Somali warlords.

While a modern state is seldom a benevolent actor, at least in the western countries it's the best of known alternatives.

In this framework violence is a key tool of the state, just like good governance, tax collection, etc.

Given that violence is a necessity, in my opinion designing military instruments in itself is not evil. It's not as good as designing new vaccines, so there is some ethical scale in the matter, but I'm damned if I can put it in concrete terms.

Are you sure it's a good idea to become less influential on the world stage?

But, there is power in organizing, or unionizing, for collective bargaining.

Well, given how unimaginative a leap the national socialists took from the national fumigation program to the gas chambers, I doubt there was much more to it than "take bug poison room tech, use it on humans".

There's no such thing as an "engineer" in the real world, and there's almost always somebody more unscrupulous, devout, or patriotic than you who knows or is willing to learn what it takes to finish the job, and school is not going to get in the way.

Added: Engineering is a practice, not an immutable identity. Even if you publicly shame all the "engineers" away from contributing to an effort you disagree with, replacing them is a matter of learning enough of the practice to complete the task. Even if we were just talking about what is in paper books today, it's hard to say that anyone could have control over who is or isn't capable of engineering efforts.

[1] https://www.amazon.com/Dream-Machine-Licklider-Revolution-Co...

This is a national budget problem. If you're in the tech world long enough, it becomes pretty clear that the only way to get Big funding is through military affiliation.

This puts a huge selective bias on what kind of technology projects actually get big funding, and further it prevents the benefits of those projects from reaching the community for years, because the military overlords demand secrecy sole use of the technology until it gets superceded.

We need to cut a huge chunk out of the military budget and give it directly to the tech sector, so that big innovative projects are actually possible without having to be military.

MIT's Lincoln Laboratory made 27% of their revenue (roughly a billion dollars) in 2016. That's one research center.

http://web.mit.edu/facts/financial.html

I believe the silicon research/production that started Silicon Valley was for military needs. Military paid for it.

First Computer of US and Britain? It was funded and created during war for military purposes.

And what about China and Russia? Will THEY stop AI research just because Google employees demand it?

I've called SV the military's generalized R&D department.

That's true, but not really in the general way you argue. Military spending on bay area R&D has been more or less insignificant for the entirety of the internet era. It exists, sure, but it's not driving meaningful revenue for any of the big players. Silicon Valley since the mid 90's has been a consumer thing only.

And even in the genesis of the valley, it was just one serendipitous application (missile guidance systems, which were willing to pay huge sums for early transistors which were literally 100x lighter than vacuum tubes).

Not only that, but when Youtube and Google block content, censor people, or otherwise prevent free speech, the common rejoinder is "they are private companies, they don't have to abide by the 1st amendment". Any company that is received taxpayer money, whether through direct subsidies, grants, partnerership, or any other avenue, ceases to be a "private company". I believe strongly that if you are truly private, and you operate entirely with your own private funds, and wholly-owned, privately purchased infrastructure, you are free to say and do what you want - silence any voice or opinion that you don't like. But once you receive public funds, in any context, along with that comes with the responsibility to the public. Despite being "the way it works" currently, this system is incredibly corrupt and logically inconsistent. In a world where the Constitution was respected, enforced and held inviolate, a company that received public funds (like Google) would not be allowed to declare themselves a private company with total autonomy on one hand, and grab countless millions in taxpayer money in the other.

"It's hard to define things, so why bother at all?" ... "Attempting to circumscribe the affects of your work is difficult, so why have a moral stance at all?" ... dancing through loaded assumptions like "duly elected" and "democracy", finally concluding with a tired crescendo of capitalist "competitive disadvantage".

I would merely counter: if we are the future, then we can't all be lazy sods, especially those of us empowering the greatest systems, information and power structures on the planet. Give a damn, it's your moral duty. Intelligent people recognize this.

This is a false history. In the 1960s the U.S. was full of young people questioning whether they should just do as they were told and be a good, dutiful employee.

Since then there has been a massive campaign to roll back what was called “Vietnam syndrome” — the idea that you should consider the morality of your actions and contributions to society, not mere legality. Hence all the passionate Hollywood WWII dramatizations, the Greatest Generation, etc., portraying war as a tough but noble effort in which we must all unquestioningly sacrifice for the greater good — de-emphasizing much of the horrendous atrocities that have been perpetrated by the U.S. military in Vietnam, Iraq, support for murderous dictatorships in Central America, Indonesia, and so on.

Noam Chomsky has written extensively on this. One essay in particular is called The Backroom Boys — a reference to the chemical engineers at Dow who developed napalm within the relative peace and equanimity of a laboratory.

> Whatever this is, and however it might be defensible in "sending a message" or whatever, it is a sure way to put a company at a competitive disadvantage while accomplishing nothing practically… If it does become that way, a new form of puritanism will hold full sway to the detriment of all.

Whatever your opinion of the anti-war activists of the 1960s, puritans they were not.

Google has users in almost all countries, and even our friends in other liberal democracies do not see the U.S. military the same way we do. Perhaps some Google users' family members have even been killed by the U.S. military. This presents a perfectly reasonable business reason (one that has nothing to do with "the personal, the moral, the legal, and the political") for Google to turn down AI drone contracts.

We are starting to see the shape of the AI driven future and it is not pretty. Autonomous drones and other robots and surveillance patrolling systems establishing strict inescapable authoritarian control.

The elite employees at the global AI leader are best suited to see the coming dangers and are sounding the alarm bells. The outlook is bleak but moral engineers are going to be one line of defense in this fight. And I hope they keep doing it.

There's a famous quote for this:

It should be noted that no ethically-trained software engineer would ever consent to write a DestroyBaghdad procedure. Basic professional ethics would instead require him to write a DestroyCity procedure, to which Baghdad could be given as a parameter.

https://en.wikiquote.org/wiki/Nathaniel_Borenstein

Of which I believe the meaning is yes, it's evil. It's handing a toddler a loaded gun sort of evil. If you DestroyBaghdad, you've limited the harm your program can do to what is specifically required by the situation. DestroyCity is easily misused in the wrong hands and should be carefully considered by ethical programmers.

Doctors solve this by disallowing unethical members of their profession to legally practice. Programmers should consider becoming an ethical profession, because depending on others in the field to do the right thing and police themselves hasn't been working out.

Do you not believe that society is only the sum of its parts? Do you not believe that the mathematics of society can be changed, the more parts of the equation object to letting their talents be used for unscrupulous goals?

I would point you towards any cultural shift in modern society, and how it began—usually, as the imbalance of classes further divides, until one class can't tolerate it any further, and uses what power they have to reset the scales. That is what is happening today, and it isn't a fluke of the short attention span of the beast of humanity. It is a conscious, concerted effort of people in this country who are tired of existing in a system devoid of morals. And to frame it as something like embracing the status quo, or becoming a puritanical society, is simply a false dichotomy.

You might want to read some history. E.g. history of the development of nuclear weapons.

This sounds remarkibly similar to the "superior orders" defense given at Nuremberg [1]

(Those who gave it were hung from the neck until dead.)

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

But I also wonder, if your employer asked you to provide some legal cover for something you found unconscionable -- like maybe draw up incorporation documents for organizing sex tourism to a place where it's both legal and likely to involve slavery -- would you subscribe to the same argument about how you should just do your job and not ask questions, how someone else will do it if you don't, and dutifully provide the legal services?

And if you think that hypothetical scenario is meaningfully different from this one, could you describe how? (I don't mean to try to back you into a rhetorical corner -- I'm genuinely interested in your response.)

This is basically the "just following orders" defense. Your argument rationalizes doing nothing.

As a moral person I think you basically have two choices: either don't work on things you think can be used for evil or if you do work on those things step up and make sure they are used responsibly.

You never know what will be the initial trigger for change, who will be inspired or whatever. Take this guy: http://www.ecns.cn/2018/01-02/286632.shtml Why should some basketball celebrity and his campaign have so much effect?

Google supporting American military may have negative effects on Google the corporation, especially considering that the rest of the world is a bigger economy than the US.

Something I would add is that a lot of people don't understand how fundamental military R&D is to the collective progression of knowledge and technology. Take almost any common technology that we use today (computers, gps, rockets, airplanes, cell phones, radios, the internet, etc) and you will find it came from military R&D and use in war.

Since AI and all its related parts are the new technological hotness, to put it mildly, it only makes sense that Google, one of the companies on the forefront of this technology, would work with the government/military to do research and find ways to apply it with their scope.

You made a point of talking about cryptography. The US government also classified crypto as munitions in order to control its usage and export: https://en.wikipedia.org/wiki/Export_of_cryptography_from_th...

This had very negative effects on cryptography in general (see the FREAK exploit: https://en.wikipedia.org/wiki/FREAK )

Also, why wouldn't this be employed to better monitor Stingray (https://en.wikipedia.org/wiki/Stingray_phone_tracker) systems?

I don't think the government (and specifically, the military) should be viewed as a non-partisan actor when it comes to technology.

This is my perspective, because I read too much scifi: I can conceive of a couple possibilities in 300 years.

1. Humanity is gone or stone-aged. Either because it failed to colonize before being wiped out by disease, because it nuked itself, or because it implemented AI in a way that got itself killed.

2. Humanity has turned peaceful, formed a global community (if not wholly, then at least nearly so regarding scientific resources), and colonized.

I draw such a sharp line because I don't see how the resources to colonize can be mustered without justifying it with an arms race unless peaceful cooperation is established. Without this, I think somebody is going to put a nuke in orbit around Mars and call themselves King/Queen of the inner solar system.

Therefore, I think we should be working towards the cultural changes that I believe necessary immediately. This is why I'm a staunch proponent of universal education, universal healthcare, gun control, etc. I think there's no better time than the present to "be in the race together" globally than now. Given climate change, death of bees, super-bacterias, and narcissists with penis-size complexes holding fingers over red buttons, I see a ticking clock.

Avoid (1) by tackling culture, is my theory, and I think that's what these googler employees are here. (I post this very much looking forward to being challenged on all points)

$700,000,000,000

Or as Obama said in the 2016 State of the Union:

"We spend more on defense than the next 8 nations COMBINED."

Keep in mind, most of that 8 are allies.

I struggle to see why we should be praising a self-fulfilling prophecy.

For the purposes of being better at waging war. I am not disagreeing with your statement, but I think leaving out the reason why is treating the military like it's "NASA with a few guns laying around" and implying that one can work on weapons systems without having any moral culpability for how they are used.

Possible reasons, using Wernher von Braun's rocket development as an example:

- He, or somebody else, would have invented rockets anyway at some point

- That 'some point' would have been much earlier if the money spent on military (research & some of the rest) would instead have been put into rockets in the first place

- even if the two mechanisms abovc were not true: I'd probably be willing to forgo the benefits of the space age if it meant undoing WW2 and the Holocaust

- In most ethical frameworks except utilitarianism, it's strongly discourage to accept some evil for some perceived greater good. Meaning: Even if rockets have saved far more lives than WW2 vanquished, participating in the project is morally dubious at best

Specifically to Google's project:

- This seems to actually invert the usual idea of the military funding basic research that later gets adapted for civilian use. They are building on the fundamental research into AI already done by Google and others, and using it very specifically to kill people. Any advances useful for other purposes would seem to be even more coincidental than usual.

- While "minimising collateral damage" i. e. civilian casualties may seem to be a worthy cause, it cannot be ignored that such advances are likely to result in greater use of the technology. Drones are actually the best example of this effect. Just look at the Obama administration's willingness to expand the use of drone strikes, which was a direct result of the technology appearing to be the lesser evil compared to manned missions on the ground.

> DoD/Pentagon involvement in artificial intelligence research shouldn't be viewed as a necessarily bad thing.

Let's be spesific: they are using current AI algorithms to track everyone's movements from airborn imagery. It will be used to kill people in drone strikes and for population control.

Another thing is to specifically choose to dedicate resources to one of those areas instead of another area. Why can't google bid on AI projects for better detection of geological landmarks in other planets? (which I'm guessing is a similar endeavor to what the article mentions - and already has ongoing research)

https://steveblank.com/secret-history/

That's a direct application of tech to killing people, not some abstract network or compiler tech. There's a big difference.

This also has the benefit of getting war averse liberals to say nice things about the large standing army and bloody foreign interventions of the US empire, most of which have failed to produce anything other than blowback and civil war across the globe.

https://archive.org/details/NoamChomsky-ScienceMilitaryFundi...

- The Global Assassination Grid: The Infrastructure and People behind Drone Killings: https://media.ccc.de/v/33c3-8425-the_global_assassination_gr...

- The Drone Papers: https://theintercept.com/drone-papers/

- Graphs, Drones & Phones: The role of social-graphs for Drones in the War on Terror: https://media.ccc.de/v/32c3-7259-graphs_drones_phones

Also, many people, including myself, consider drone strikes to be state-sponsored terrorism. Just because some part of the military may be justifiable, does not mean all of it is.

The Google workers urging their CEO to pull out of Pentagon projects are the ones who do not have such short memories as to be distracted from the blatant deceptions and disasters carried out in recent history, with no sign of relenting.

What the google employees need to do is push for a civilian government research and science funding effort, like ARPA was back in the day. Unfortunately that sort of funding is politically dead, while military funding is always in style.

Reminder, if you work in tech you are collecting blood money because everything we use was initially funded by the military for military application. If you are going to attack Google for this and don't leave the industry you are a hypocrite

And what do we have to show for it? Wikipedia is great and Youtube is entertaining, but the rest of the internet is horrible. I say that as a software developer.

Due to the wonderful internet our children now have attention spans far below the time a typical class lesson takes. The smartest brains on Earth are not building spaceships or curing disease, but rather implementing systems to get you to click on ads. Our bodies and habits are being tracked, our opinions manipulated, and our privacy is being handed of willingfully for a chat and photo-sharing platform.

Military tech _often_ has implications far beyond what the original scientists and engineers intended. Sometimes for the best. But as you allude AI to the internet, don't forget how much damage the internet is doing to us even as we enjoy it and profit from it.

I think DoD/Pentagon put up the funding for what later became Google maps. Not to mention the GPS constellation we all enjoy using.

Its a shame progress has to be made through these channels but it has proved very effective in the past.

We need weapons to fight an enemy like this, with help from US technology companies.

also it is not surprising that these things are invented by the dod when you're spending ~600-700 bilion a year on warfare and intelligence operations.

That last sentence of yours rubs me kinda the wrong way; that is a dangerious premise that could set us back ~30 years when the cold war and the arms race was well and alive, but this time we'd have robots and AI

Are you basing this statement on a general feeling or on a specific data-based metric? The military industrial complex greatly benefits from this common sentiment, but by most all quantitative measurements the world continues growing slowly safer, average lifespans increase, etc. While years-long aberrations exist (e.g. civil wars), the overall trend is very clearly antithetical to your statement.

What are you advocating? Arming governments with autonomous killer drones in order to "protect democracy"? Because the world is a "dangerous place"? I don't even...

Nope. Not even in the slightest.

https://en.wikipedia.org/wiki/File:Waves_of_democracy.png

https://en.wikipedia.org/wiki/History_of_democracy#Waves_of_...

Pacifism isn't viable when some of the biggest countries in the world are run by totalitarian regimes. Objections to participating in the military industrial complex are a luxury of countries verging on global irrelevance. The proclivities of the West to engage in self-flagellation over these issues are well known to and exploited by less liberal regimes. If only it were not so, but I don't know what the alternative is, just as the physicists in the Manhattan project saw no alternative.

After all, I would not be here, and my job would not exist, and the freedom I have to do this work would not exist, if soldiers like my grandfather hadn't stepped up to Hitler and Stalin and Mao and say "No, you're wrong, and we're willing to use violence to protect our way of life". Just my 2c but I definitely respect the ultimate sacrifice that my grandparents generation gave to create this quite nifty Liberty bubble that these few generations get to play in.

Keeping "free" nations safe and prosperous is as fine a goal as any.

Boy they pay great though...

These smug generalizations, if you did not vote for X, you are racist! If you are not for A, you are for Genocide! are not going to win you over people that are not already in your camp.

You come across as wanting that sweet sv $ and benefits without having to make any moral choices - if you got to work for CNI company's you know what the deal is.

And before you ask yes I did face that at my first job when I analysed some high speed film from one experiment and though oh I know what that "objct" is and my office mates said yes that's probably REDACTED.

I don't agree with this. Many open source projects have been used for sinister means (not just by governments). But does that make those contributors complicit? Should those contributors be safe-guarding against all possible unintended use cases?

It seems like you're painting all military efforts and funding in evil terms. But these efforts are directed by many different people, each with their own goals and agendas. Some of those efforts are intended for the betterment of humanity, while other efforts are at the detriment of humanity (at least in my opinion).

I worked for several years at a large defense contractor. The majority of our funding came from military contracts. I worked on a project intended to help detect and disarm explosive devices. And I worked next to a team that was trying to rapidly detect infectious diseases. We all thought we were doing work that would benefit society. But virtually any technology can be used for malicious purposes. Does that make us complicit if those technologies are ever used in malicious ways? Do the evil actions of some people in an organization make the entire organization evil? Does that mean that all military efforts are unjust?

Technology is neutral - who it's used by and how is not.

I don't care what you build, if it has the goal of improving a workflow or automating a process then it will de-facto support one organization over another.

People refusing to work on military technology isn't "appearing moral". It's just "moral", at least if you consider war to be generally amoral.

People here and elsewhere have been throwing the term "virtue signalling" around a lot in the last year or so. This is a perfect counterexample: Google employees speaking out against military tech are shouldering the risk of appearing disloyal to their employer. If Google were to cut their military ties, they would forgo whatever benefit they previously sought in that relationship.

I don’t really get why anyone would want drones to be less accurate, but maybe that just means I’m just a no-good warmongering fool, right?

> They will not have any incentive to rethink the sale of user data to advertisers

Is it not possible that one will lead to the other? We're seeing the beginnings of a real examination of the power Silicon Valley giants have. Recent revelations about Facebook, amongst others, shows that awareness of advertising profiles etc. is increasing as well.

So while it's great these protestors want to protect us from AI, those in China, Russia, India, Iran, will endure a greater threat from AI and it behooves us to be at the forefront lest we fall behind.

In other words a moratorium can only work if all participants agree and participate. Sort of like the anti nuclear proliferation treaty.

Note : I do not agree with the petition, but the question was too tempting.

More seriously: the US has nuclear weapons, rendering them safe from any direct confrontation with enemies challenging them technologically.

They have a conventional military 10x the size needed for the sort of intervention that actually succeeded in the past (the Balkans, the first Gulf war). The rather unsuccessful wars (Afghanistan, Iraq) were lost not for technological reasons or any other dimension of military might, but because the US military sucks at convincing people of the benefits of democracy.

Want to win the next ten wars? Put the money into Hollywood blockbusters and the world's best Halal fast food chain.

A lot of Silicon Valley's justification for its immoral actions rests on the _belief_ of the inevitability of outcomes. It's ok to monopolize and centralize everything, discarding the lives of people who are negatively impacted because it's inevitable. It's ok to make robots able to kill people better because it's inevitable, etc.

The thing is, it's a _belief_ that this is inevitable — not a proven fact of nature. And people who don't believe in it will logically withdraw their cooperation from building that future.

Both sides are 'proving' their point of view. If enough people object, then it wasn't inevitable after all. If enough people work to bring us into that future, maybe it was. But thats the struggle and moral choice we face.

A) not be developed without support from Google, which is categorically untrue; The DoD might simply find it more difficult to acquire top talent if that talent is opposed to the work on philosophical grounds.

And B) be necessary for the US military to remain the most advanced and/or powerful in 20 years, which is also exceptionally doubtful - unless you believe ML to have such effective wartime applications as to render a nearly threefold larger military budget (compared to China, the next largest) moot.

Nukes are enough to stop any war from happening between any of the most powerful countries in the world.

What else do we even need?

The various factions in the various civil wars around the world can kill themselves in their own war, and we shouldn't be a part of it killing even more people.

To put a number on it: I am absolutely certain that Facebook would be willing right now to sacrifice 25% of their revenue for a bump in their reputation (among developers only) that puts them at Google's level.

(In the same vein, Google's NetPAC has been quietly donating the max to Republicans like Paul Ryan and Steve Scalise, while most of their engineers openly oppose them.)

And you want to make the program "more efficient."

This is the gulf which I cannot seem to bridge.

Extra-judicial killings of "terrorists" in "bad places" using flying killer robots is just batshit insane to me. I cannot for the life of me understand how that conversation would go with, e.g. Thomas Jefferson or Alexander Hamilton. Or really anyone who believes in human rights.

Then “AI” should be the least of your worries. Not everything is a software problem.

Their only intention here is to scale up their drone programs.

This is nonsense, Google has plenty of revenue and other non-defense avenues for growth.

Companies are not beholden to chase every last dollar, especially if it could alienate their workforce.

I don't have the same anti-defense leanings as others around here, but to say Google has no choice is ridiculous.

Google probably couldn't (and shouldn't IMO) stop the DoD using Android phones if they want to, but they can certainly refuse to work with the defense sector.

Perhaps not, but that doesn't necessarily imply that Google can't pick and choose what technology it sells or provides to the military. They can, for example, serve the administrative needs of the military (e.g. Google Apps), but not necessarily provide technology that directly improves their weaponry.

Take, for example, RADAR. Should those that developed RADAR, during or just prior to WWII, and abstained because of its possible wartime use? Wikipedia says "In the 1934–1939 period, eight nations developed independently, and in great secrecy, systems of this type"[1], so those that were left out faced a real danger in the coming years if they didn't have something to compete. Yet, RADAR has been used to great effect in many weapon systems and by many countries since WWII in ways that that most would agree are horrible. Then again, it's made commercial air travel much safer than we could have hoped to achieve at this scale without RADAR.

It's a complex situation, and I think it deserves a complex discussion. Distilling it to "working with the government is bad because I don't like what the government is doing" doesn't really do it justice.

1: https://en.wikipedia.org/wiki/History_of_radar

Or is there something specific to Google's Maven project that makes it shameful, while other types of military commissioned projects might not be?

It's also possible that when a weapon is deemed technically capable, people will be more comfortable with broad use of weapons and with expanding lists of individuals targeted for extrajudicial killing.

You do not need advanced technology to build simple rockets with simple warheads and wreak havoc. These rockets are guided by nothing better than pointing and shooting. We see this in use in various paramilitaries around the world with little concern for civilian casualties: in Gaza, in Syria, in South Lebanon, in Yemen. If the West was similarly limited to such crude munitions then the civilian cost to destroying legitimate military targets would be enormous. Better guided munitions reduce civilian casualties and as such working on making munitions better-guided is indeed ethical.

If you think that not giving generals the tools to better avoid civilian casualties means they will simply not wage war in ways that endanger civilians, well history is pretty clear that this is quite erroneous. Giving generals the tools to wage war more ethically allows them to make decisions with fewer compromises e.g. civilian casualties.

This is not even to mention work on purely defensive munitions, e.g. Iron Dome, which is inherently ethical.

And if your opposition is rooted in a sense of pacifism, please, freedom is not free, it's not just a slogan. Anarchies (i.e. the lack of a sovereign government with a military defending territory) and their power vacuums result in warlords and the end of civil society. And that Malthusian outcome is clearly less ethical than almost any other. That's just human nature. Defense of civil society requires a military. Ideally, that military will act as ethically as possible while achieving its primary goal which is the stability (in face of external threats) necessary for civil society to flourish. There are two ways to make the military ethical: a) give it ethical tools so that it won't be forced to use less ethical means for achieving its raison d'être, b) set ethical policy. As sad that it may be that we have little control over the latter, as engineers we can have control over the former.

Our military doesn't get to pick the fights. They get to fight the fights.

http://nuclearfiles.org/menu/key-issues/nuclear-weapons/hist...

From the Soviet side, Andrei Sakarov.

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

When I was running a robot company, I made the decision to not do military projects (mostly for personal reasons) and it helped me hire some great people that I couldn't have otherwise.

Why go down this path at all? Google doesn't need to be a war contractor (I think "defense contractor" is a misnomer these days - we all know what this tech is used for: bloody violent and usually unfair/illegal war).

Google should avoid becoming a sponsor of death in the world. What happened to "changing the world" (for the better, I assume)? Is feeding the American war machine changing the world for the better? Really?

Google hasn't been doing that, ever. That's really the point of the story - the agency of the people who work at Google.

The fact that you are outing yourself (now and in your HN profile) is very unlike any other "spook" I know.

It's laughable that you say your " background is all out there for people to see".

Common misconception, and pointless to even bring it up. Because while "Don't be evil" is in their code of conduct, they regularly do evil.

Holy poopers, no. Drone strikes are not like premarital sex. The fact that we're _here_ (that being autonomous killer robots firing missiles from the sky to kill "bad guys") and no one bats an eyelash, means we're already in some kind of massive fail scenario.

We could, for example, go back to the rule of law. You know, before we believed that every bad guy out there was going to end civilization and deprive us of our "freedom." Before we decided blowing up wedding parties in Pakistan was such a great use of our time and money, and every Allahu-Ackbar-kaboom moment is a pervasive existential threat.

A petition by 3000 employees (or, to be honest, 30) is enough to claim that it was in the company's interest to forgo this revenue to improve their standing among potential hires. Remember that tech companies live and die by the quality of their workforce.

Moreover, the idea that companies always have to act according to strict profit motivations is a myth. Just look at the billions corporations spend each year on charitable causes.

Yes, you could claim that these donations are done for PR purposes. But that would render your argument meaningless, because I can reframe any such altruistic actions in terms of PR. I have also personally witnessed many decisions that cost money but served some greater good and were never publicised.

https://techworkerscoalition.org/

But this seems an amoral thing to do, to me! A small group of unelected, unaccountable and non-diverse people abusing their relative power to try to force a democratically elected government or private company to do something? Why would anyone want that?

Wow, we really are insulated from our military today, aren’t we? Yes, the US government is still at war in Afghanistan and with ISIS.

In this case what they seem to be doing is allowing the military to use some of the AI tools on Google cloud which are also available for general use. This is a much grayer area - it is one thing to not develop something specifically for the military, it is another to say that we will not provide the military something we provide to the general public.

"If we don't employ child labour, somebody else will! It should be us, so we can at least make sure our child-slaves get fed once a day!"

The fact of the matter is that Google has rare talent in AI. There isn't another company with their kind of expertise waiting in line behind them, and even if there were, sometimes you should keep your hands clean.

Even Sun Tzu notes that warfare is not total.

[1] https://marchforourlives.com/how-we-save-lives/

'Fundamental rights' (under US law) I think is defensible as fundamental is treated as an alias for constitutional so guns are in.

'Fundamental human rights' to me implies broad and fundamental applicability, i.e. the UN declarations and covenants on human rights, which don't iterate guns as a right.

Case in point, when non-US countries talk about human rights, none of them are talking about guns.

Suppose you decrease error rate by 10% (optimistic), but increase volume by 20% (probably underestimated). You'll get an 8% increase in innocent deaths.

And tolerating the opinions of others.