Nothing but respect to TurnTrout for taking an action like this. The world needs more smart people who are willing to stand for what they feel is right, despite the pressures otherwise. Without that occurring more, our species is going to lose many impactful prisoner's dilemmas coming these next two decades.
This raises my respect for AI researchers a little bit too. I have often felt that the entire industry is pretty tainted to the core, and for better or worse that colors my opinion of the researchers.
Maybe I'm in the minority, but I thought it was gross to download pirated art for a student project when I was at Berkeley years ago. So it has been really sad to witness many of the most brilliant minds of this generation answering the siren song of disrespecting the collective effort of others to extract and resell residual value.
I'd guess TurnTrout doesn't agree on that framing, otherwise he probably would not have been at Deep Mind. But clearly he and I agree on other ethical positions; I am nothing but glad to see him stick to his principles here.
> I'd guess TurnTrout doesn't agree on that framing, otherwise he probably would not have been at Deep Mind. But clearly he and I agree on other ethical positions; I am nothing but glad to see him stick to his principles here.
FWIW I agree that creators should be compensated (but evidently it wasn't a deal-breaker for me joining). I think it's bad how little that has happened.
I joined GDM to work on AGI safety to reduce existential risk from AI. I consciously avoided work that would improve the raw capabilities of Gemini. When I joined, I made a trade-off and it's valid to disagree with my decision.
As your writing seems to allude to, life is an impossibly complicated series of trade-offs made with imperfect information, so I definitely wouldn't consider joining an AI frontier lab to be a priori bad. Even if I have some qualms with how the models are built.
If anything I was just trying to point out that, even with people we might disagree with, we deserve to show them recognition and respect when they make moves we do agree with.
But I doubt we disagree on much. Your writing is great, and it makes me sad to see this thread somehow disappear from the HN front page so fast. Either it triggered some internal flamewar detector--but without a flamewar--or someone didn't find it convenient.
I wish you the best of luck with your next endeavors, and look forward to seeing the value you add to this world.
Hmm, links for what? My student project? The decisions that face humanity that are pretty clearly modellable as prisoner's dilemmas?
Otherwise not sure what I could cite--I would assume most all on this forum know that AI is trained on the works of other people, without their permission to do so. I guess you could disagree with my framing, but I wouldn't think this requires a citation.
I think maybe my writing wasn't clear, and it sounded like I was referrencing some well known thing that happened at UC Berkeley. I have edited it to read more cleanly!
It wasn't clear to me it was _your_ project nor who pirated it. I was under the impression it was a well known scandal from your original, unedited comment.
Apologies if my comment sounded hostile; I was just asking for a clarification/more information on it.
Props to the author. I left Microsoft due to their work with Israel to spy on Palestinians and record all of their phone conversations (in addition to other IDF collaborations). They ended up walking some of it back, but Satya was complicit.
I salute you for your morality. If most people did this, those companies would change their position very quickly. After all, they only think about money.
Thank you but I'm no saint. I joined Microsoft knowing that they were incompatible with my values. I just told myself that I wouldn't work on those types of projects and try to stick to my values. I told myself that I could just be "professional" and keep my head down.
But, as I learned more I was filled with a deep sadness and shame. It was tearing me apart. Every day I got up to go to work and was getting sick to my stomach.
Now I just feel lost. I spent my life building a career in this industry and it feels like I have nowhere else to turn. Microsoft is not the only company actively complicit with horrible crimes and human rights violations. I've loved technology since as far back as I can remember, but looking back with the knowledge I have now, I just feel dirty. I honestly don't even know what to do at this point, I do have a job now but when I look at who funded us, it's every bit as bad as working at Microsoft. I feel like I need to walk away from it all and start a food truck or something.
It's more rare, and definitely less lucrative, but there are opportunities in software engineering for non-profits, open-source, and smaller non-tech companies. If you've made your blood money and don't need top n% pay, it's something worth checking out.
I haven't finished reading, but I should note for the group that the author doesn't represent the Anthropic situation accurately. Emil Michael, the undersecretary of war/defense for research and engineering went on the All-In podcast (1) and explained, for quite some time, exactly what happened. You should go listen to it (2). In essence, in the negotiations, Dario kept coming back saying "well, if you need it, call us, we can redline things as needed". This happened over and over and over. Emil's point was that a major conflict, if it were to occur, might happen on the 30 minute clock of an ICBM, and all due respect to Dario, the national security apparatus just doesn't have an allowance in that 30 minutes to call him for a redline. Emil felt Dario had demonstrated he wanted ultimate control via line item veto, and was willing to trade up to and including the survival of the nation for that veto. And a government cannot be expected to pay for that sort of behavior from an entity domiciled in their jurisdiction.
Now, Dario is going to win a decent slice of the economic pie. But as an military acquisitions matter, I gotta say, I have to agree with the undersecretary's position here, and yeah, it makes sense to document a company's undesirable behavior, and in certain circumstances push that information to others. Not the first time the government has found a company under contract acting in a way that appears to be counterproductive to the government's obligations; there's a whole database full of this stuff (3).
1) Without a doubt, All In is a friendly crowd for Emil, but I think that actually made it easier for him to get more nuanced facts out in this case because he wasn't spending a lot of time defending malignant attacks.
I cannot think of a less trustworthy source of information than Emil Michael on the All-In podcast. Why on Earth would you take his statements at face value? I grant that nobody in this situation is totally trustworthy, but if it's his word against Dario's I'm believing Dario 10 times out of 10, and it's astonishing and dismaying that anyone might do otherwise.
Keep in mind that Emil Michael's job at Uber was to slander and destroy their political enemies. What makes you think he hasn't continued exactly as before?
Even at zero dollars and a billion dollars they feel rare. Honestly I am surprised how many tech people still work for Facebooks and Googles of the world when they actively have used their platforms against most of their values. These are skilled people that can still make a living elsewhere.
I Think it’s because when you have enough money to never have to work again and you can move to some place nice, you can isolate yourself and your family from the rest of the world, like most rich people do.
So that’s what, 4-5 years at a society destroying tech co in exchange for lifelong freedom away from the people you have actively made life worse or impossible for.
It seems many people can live with that and in fact will jump at a chance to do it.
And there are others who left years ago and warned us. Back-patting isn't all or nothing, and I think it's reasonable to praise those who left when they saw the writing on the wall more than those who leave only once they see their company materially supporting state killings.
I did not know about TurnTrout till now and thanks to HN community now I know. I am beyond impressed as to how he lives by his principles. I don't think I have this kind of courgae or confidence and thankfully I don't work for Google or have to quit it. On the other hand we have smart people who have choose to work for companies like Palantir and are proud of it.
It is not about courage or confidence, it is more a calling, where few are called and that often turns out to be just one person, or so it seems, since everyone finds their excuse to not take a stand against the oppressor and for the oppressed.
The author will have experienced every logical fallacy and TurnTrout has done well to document his efforts, rather than just the crimes of his adversaries.
A calling of conscience is hard to articulate initially, it is conscientious objection, to the war machine and the system of finance that necessitates the empire that needs the war machine.
The author starts off with a high degree of authority in that he actually worked for Google Deepmind, however, nobody will listen to him, so he has to seek higher authority to carry the truth forward. But he could have gone to the pope, anyone short of Jesus Christ (freshly teleported back) and it will be a no, everywhere.
Logic and reason does not help when insanity and money have taken over, which tends to happen during wartime.
The conscientious objector may not believe in god, however, they will consider themselves answerable, 'vain' enough to care about their legacy. There is that desire not to be found out, doing the 'devil's work', generally in workplaces where everyone is glad to get on with the 'devil's work'.
Why do so few people make a stand? Why didn't hordes of five-eyes people walk out the day after Snowden did his good deeds?
It varies by individual, however, the author is vegan, which means he has already 'dared to be Daniel, dared to stand alone', albeit only in a lunch queue in a meat eating world. This requires living according to principles, and serves as target practice for war-time conscientious objection.
Also important is a certain level of independence, the guy with a mortgage and a couple of kids, underwater on the car, with maxed out credit cards cannot conscientiously object. Nor can the guy counting down the last few years to retirement, and then the very young lack the articulation.
Only a few have the 'warchest' to embark on an open-ended single-person campaign to confront power with the truth. We owe a lot to these people, particularly the 'failed whistleblowers' that don't make it to 'whistleblower' status, because the media then makes the story about their situation, not what they conscientiously objected to. Props and respect to the author for documenting his journey and taking the first step.
Note the 'first step' is classic 'hero's journey', where the call is initially rejected, but then a journey into the extraordinary world is made, where the challenge is to bring back what is good from the extraordinary world to the ordinary world, for the benefit of all.
Respect for sticking to your beliefs. Just out of curiosity though why do people not want smart AI weapons? I would much rather have an onboard AI that can discriminate between unarmed civilians and military assets, seems irresponsible to not... Is a dumb sea mine that blows up everything somehow better than a smart sea mine that knows to not blow up sometimes?
"You're absolutely right, that wasn't a military target—it was actually a girls school. It won't happen again!"
I realize that's not a great argument and was definitely tongue-in-cheek, but given there's still a lot of debate about the accuracy of AI for far more mundane tasks, my personal perspective is that until we have LLMs and such that are truly, demonstrably far more accurate than humans, with true reasoning and judgement capabilities, they don't belong where lives are at stake.
I wouldn't want an LLM-underpinned machine running anesthesia during a surgery; why would I want an LLM-underpinned military apparatus that is deciding the lives of far more? I wouldn't, not in their current state.
In a hypothetical future where we truly trust incredibly smart AIs or LLMs or whatever "smart" technology it is for driving weaponry, okay - if it's truly necessary; I abhor war and the death and destruction wrought by it.
In my mind, though - even if we get to that future where there's some vastly superior technology to the LLMs we have today, which can judge and reason, then I'll have a bunch of other questions, like understanding the motivations of said technology, because I suspect it'll be something much closer to AGI, and that opens a whole separate can of philosophical worms.
Not only that but an AI could follow any orders it is given (whether it hallucinates that's a different story) whereas humans may oppose. The friction of orders having to go through a chain of command through its completion is a feature not a bug and that adds responsibility. AI can take no responsibility by itself and with 'moving fast break things smart weapons' that responsibility can be further shoved under the rug.
> "You're absolutely right, that wasn't a military target—it was actually a girls school. It won't happen again!"
What if it was the target though? AI may be more capable than we're giving it credit for (especially the AI accessible to the US and other governments). The attack on the girls' school coincided with Purim and I don't believe it was a mistake. I think it was the opening salvo by a radically religious Zionists (Christian and Jewish).
I don't think the biggest problem with AI weapons is that they make "mistakes", I think the biggest problem is that they allow people who want to kill civilians the ability to accurately do just that.
So you would prefer weapons that cannot reason at all?
> "You're absolutely right, that wasn't a military target—it was actually a girls school. It won't happen again!"
Most likely this event happened due to a bad targeting system that wasn't smart enough, if it had a better llm underpinning it (assuming it had any) maybe those lives would have been saved. More reason for more smart people like the author to work on these systems.
> So you would prefer weapons that cannot reason at all?
Um, yes? It's bad enough humans are murdering each other. At least a human can in theory be held accountable for pulling the trigger. The last thing we need is an unaccountable ralph loop reasoning about which schools and churches to bomb every time it wakes up.
> if it had a better llm underpinning it
Ah yes, the "LLMs are intelligent, you're just not using the newest fanciest model" fallacy. This time with innocent lives on the line. If only we used ChatGPT-8.9 instead of 8.8, those poor kids would still be alive today.
Mainly I don't want them because I don't want a machine instantly making life or death judgements. It might be fine if it mostly informs a human who ultimately pulls the trigger, but having that failsafe is important in keeping machines from going on killing sprees.
I also want people to be held accountable when they do unjustified killings. AI weapons make it FAR too easy to simply pass off a killing as a "woopsie doodle." It's just not acceptable to say "The algorithm made a mistake, version 23 will do better".
I don't have a problem with the AI providing additional information to it's user, but when that's incorporated into a weapon it's a short distance from that to completely automating the killing.
Responsibility will always have to lie with the user, but if the weapons are more advanced and safer then why is that bad? I like that autonomous vehicles like Waymo are 10x safer then human drivers, even if a 'Machine' is making decisions.
That is a great point and one I never considered. The folks doing the murdering shouldn't have their jobs be easier. They aren't safer for the person getting killed.
> Responsibility will always have to lie with the user
Being fully autonomous makes it hard to identify exactly who that user is and is easy to dilute responsibility. Perhaps someone was added to a kill list by mistake. Maybe some internet hi-jinks tricked an AI into falsely identifying someone as a kill target. Perhaps it's the case that someone was in the 5% of a 95% confidence of identification. I'm not a fan of putting killing into the hands of something known to get things wrong 1 in 20 (95%), or 1 in 100 (99%), or 1 in 1000 (99.9%).
> but if the weapons are more advanced and safer then why is that bad?
It's yet to be proven that they are "safer" as they become more advanced. There's also a question of "safer to who". It's technically "safer" for a soldier to shoot first and ask questions later, it's obviously not safer for the villagers.
False identification, which is an absolute part of AI, doesn't make these weapons safer for anyone.
> I like that autonomous vehicles like Waymo are 10x safer then human drivers, even if a 'Machine' is making decisions.
Waymo has the reverse bias. If anyone dies as a result of waymo it's gone horribly wrong.
AI weapons are designed exactly to kill, if they don't kill when they should something has gone wrong.
A good thought experiment for this is to replace guards at jewellery stores or banks with a robot. Which one would you feel more safe? Do you trust that the robot won't randomly shoot you if you act funny on the street?
> Just out of curiosity though why do people not want smart AI weapons?
Because the minute you defer decision making to a machine you eliminate human accountability.
Any time the robot is used to do something illegal or immoral, you can just say "it was a bug" or "that's not what I really meant". It enables plausible deniability.
We see this every time AI is allowed to make decisions. That robot promoted violent extremism in your feed? It misidentified a black man? It disproportionately chose men in hiring decisions? shrug Its the algorithm. Who can argue with the algorithm?
Likewise any attempt to codify liability is typically avoided at all costs because the purpose of those systems is what they do.
> I would much rather have an onboard AI that can discriminate between unarmed civilians and military assets,
Has this been proven to actually occur? There have been lots of reports that similar systems in Israel that use AI, struck hundreds of targets that were civilian. They did not discriminate.
It’s true that humans are unreliable, even if we could guarantee a consistent set of moral and ethical standards as well as the appropriate tools and authority to enforce them at all times. We do, however, have to deal with 100% of the implications, one way or another. Machines don’t. Vasily Arkhipov and Stanislav Petrov had to consider the world they would condemn their children to if they pushed the button, the missiles didn’t. These are imperfect devices made by imperfect beings, whose rules and goals can be changed more or less at will by whomever wields them. Offloading the responsibility of pulling a trigger or trampling someone’s legal or human rights to such a machine while insulating those in command from consequences threatens to make things worse at a speed and scale I don’t think we’ve seen before in all of our dark history. Meanwhile, it further concentrates immense and unchecked power in the hands of a relative scant handful of elites, all of whom exhibit the lack of ethical and moral fiber we were talking about to begin with.
Teaching something or someone to pull a trigger is trivial compared to teaching them not to. I have no reason to believe that automated weapons and surveillance will be more reliable than our world leaders or that the world will be safer with them in it.
A regime can only be as oppressive as the people following its orders. There are only so many protestors you can order your army to kill until your troops realize they are killing their own friends and family, at that point they either desert or turn their guns against you. This puts a check on what even the most evil tyrants can do. Not a perfect one, for sure, but still better than nothing.
Technology changes this by reducing the human involvement needed in killing. The Nazi gas chambers were a historical example — it turned out mowing down innocent civilians with machine guns was too mentally taxing for even devoted Nazis, so they invented a bloodless industrialized process where they can kill with greater efficiency while also sheltering the killer from the immediate effects of his actions. A win win for the Nazis, not so much for its victims.
Autonomous weapons are dangerous for this exact reason. They bypass individual moral decisions entirely. They are the perfect "just following orders" machine. They allow a small concentration of powerful people to impose force upon the populace without ever having to fear their own soldiers. It tips the balance of power massively in favor of those already powerful.
>Just out of curiosity though why do people not want smart AI weapons?
When the decision to kill another human being is made that should be in the hands of a directly accountable other human being, not an unaccountable machine developed in the basement of a private corporation.
And mines, both dumb and smart, in particular anti-personell mines are banned by the Ottawa treaty ratified by 162 countries. It's exactly the autonomous and fundamentally uncontrollable nature of mines, not just that they're dumb, that has produced countless of casualties long after wars were over. Can you tell me that millions of autonomous loitering munitions are not going to end up exactly like those mines still blowing legs off people decades after conflicts are over? And who is responsible then?
no they were banned because they're persistent. That's also the US position despite not being signee to the treaty. Making autonomous weapons intelligent only makes the issue worse because at least expertise was a limiting factor for most rogue actors. Do you know what will happen to millions of autonomous drones when the Russia Ukraine war draws down? Which hands they will end up in? You've created the capacity to kill people 2k kilometers away for a thousand bucks on a chip.
And when those inevitably backfire and start blowing innocent people up, are the CEOs going to prison, the politicians who ordered their production originally? No, nobody is going to claim they're responsible, and that's why it's insane to build a weapon that does not have a name, time, place and direct order associated with it.
Even the US military is already incapable of taking responsibility of blowing a school with 150 kids up. You want to hand these people systems that are unaccountable by definition, that are smarter than they are, when they already struggle to understand the consequences of their own decisions?
They aren’t training smart AI weapons to discriminate between civilians and soldiers. Even so, what happens when the soldiers are disguised as civilians? Or when the enemy forces civilians to serve or be a decoy?
> It is also the story of something larger: how powerful people and institutions failed, one after another, to keep their AI ethics promises in the face of pressure.
I've worked at Google. These weren't promises, they were deceptions.
Google, like many companies, maintains the focus and motivation of their workforce via a stream of carefully orchestrated disinformation and half-truths. One of those half-truths is "We listen to our employees". They listen when it has no business consequence and they can tout themselves as being "good" or on the side of the workers. They don't whenever the politics or economics of the situation presents even the slightest resistance.
Kudos to the author for fighting for change. But the dark patterns in Google run too deep for them to be uprooted by a polite exchange of white papers and emails.
We know now that Google is a sycophantic company whose DEI initiatives were all fake and dropped once Trump got elected.
This somewhat naive initiative was bound to fail. The good news is that the AI military products won't work, except perhaps for blowing up a girls' school.
Here are CEOs falling over themselves to support Hassabis' regulatory capture proposal:
Come on, ALL companies are psychopaths in their behavior.
Their only objective is to make money to shareholders. So anything that gets in their way is fair game. They are relentless optimizers.
After all, most people holding GOOG, MSFT, META, or others don't care how I happens, they just want their 401k fund going up. It's like that by design.
Search for Ukraine mid-range strike capabilities in the last 1-2 months. It's using homegrown Ai systems to great effect, re: Crimea is cutoff to the point gas is heavily rationed or not sold to civilians at all.
It's almost like Anduril has little to no background in building military soft/hardware [0][1]. The modern day charlatans are alive and living very well.
Regardless of whether AI is involved, many people simply don't want to work on weapons systems, or on non-weapons technologies that are intended to become part of a weapons system.
Not really, it is fairly common belief. And American war system is not in the pro-democracy, defense only or whatever else business anyway. It is in pro-warcrime, pro-stupid-wars, pro-dictators business.
I think there's some nuance here that a lot of these convos dont make plain - are we talking about just using 'ai' in the military broadly, or are we specifically talking about fully autonomous weapons systems?
In the former case, I would agree with you completely, I havent heard any arguments beyond 'I dont like working on military stuff'. But if we're talking fully autonomous weapons, that's a different story. And further muddying the waters is the fact that the former is obviously a step along the path to the latter.
AI helps people scale. If being "on target" means that you find people when they're eating dinner so you can kill their entire family, then using AI to scale that up is a terrible thing. That's exactly what Israel did with Lavender. They literally had a kill list called "Where's daddy?".
AI is the end of the funnel and I actually think the mass surveillance at the beginning of the funnel is a bigger issue. The concentration of power is the problem. If you or I could defend ourselves against state level actors with AI weapons, then they would be a good thing (imho). AI and mass surveillance at the state level create a power imbalance that is a major threat to human rights and should be resisted as much as possible.
When folks who live around me vote for things I find unethical, then I don't feel bound by those decisions.
Also, your position on weapons development is premised on the idea that at lease some of the folks developing the weapons are on your side and always will be.
> For better or worse millions of Americans voted for the guy doing the deportations.
Uh, the idea of democracy is not that voting reveals universal preferences or something. It is ok to disagree with whoever gets elected, and continue working towards an alternative you believe to be better. In fact, democracy depends on that; otherwise, why not have a single election with no term limits? There is supposed to be ongoing difference of opinion.
> I also find it difficult to reconcile not using AI for weapons.
You can't reduce "AI for autonomous lethal weapons" to "AI for weapons".
It does appear to be anti-democratic given everyone knew Trump’s platform when he was voted in. However the great unifying fact of politics is that everyone only really believes in democracy when they win.
This idea that once a party gain a little more votes, people who oppose it should cease to exist and cease push for own values ... somehow applies only to the right.
When center or left wins, somehow, magically, same logic dont apply.
> Plus, a pledge is only worth the credibility behind it. When someone signs “I will not support the development of lethal autonomous weapons,” then stays while their company sells unrestricted AI to a military that wants exactly that, they teach every counterparty a lesson: these safety people will not act, even at their own brightest line. The next commitment they make is worth less. Eventually it’s worth nothing.
This is why I haven’t taken the internally employed (and visibly public) AI safety people seriously for years: they always back down, because they value that fat paycheck and fiscal rewards and associated status over their proclaimed ethical values, red lines, and boundaries. Their actions betray their every word.
That’s also why I push that we remember these names, the ones who fought back as well as those who lied through their teeth. We should remember everyone willing to harm others to protect themselves, and hold them accountable for their misdeeds. In an ideal world, people like Jeff and Sundar and Stuart would be blackballed so hard they’d never be able to get so much as a help desk job ever again, nevermind any position of leadership. It’s why I proclaim them to be monsters, giddily sacrificing the anonymous other for personal vainglory and the avoidance of standing on any sort of principle.
The OP did everything right and demonstrably proved the entire apparatus is rotten to its very core, to the point of infesting and poisoning the very entities professing to stop its worst excesses (IASEAI). Why we give these ghouls the slightest bit of credibility anymore is beyond me.
> Even if Google had adopted your Framework, the Pentagon would have refused
> I agree. xAI would still have given over their AI. But if Google had given signs of independence earlier, it could perhaps have built a coalition with OpenAI and Anthropic.
This is like saying that Google and Apple might build a coalition to prevent App Store regulation. These are competitors, they all see moral flexibility as an advantage. They're not going to take a moralist stance if their federal protection is predicated on federal cooperation. All of these FAANG businesses have already bent the knee in anticipation of this, xAI is just riding the coattails of the federal quid-pro-quo.
When you go to work at a megacorp, you're always leaving your ethics at the door. Yes, there's an attractive pie-in-the-sky fantasy that Apple does care about human rights, or that Google isn't evil, but they're always just lies. I disagree with a lot of GCP's customers, but I'm still shocked that a DeepMind employee would make it this far in the career pipeline before seeing the soylent green get made.
Competitors often wish they didn't have to compete.
For example, a cartel is "a group of independent market participants who collaborate with each other and avoid competition in order to improve profits and dominate the market"
It's something that could happen in this world, through the collective action of employees, that corporate strategy can be changed.
> through the collective action of employees, that corporate strategy can be changed.
I just don't agree with this. I'd like to, but the corporate strategy is not being set by rank-and-file engineers. Sam Altman, Dario Amodei, Elon Musk - these are the executives that are steering the ship. You can change the culture internally and cause a lot of strife, but typically you don't have control over the strategy. If the executive says you're giving API access to the NSA, then that's what is going to happen.
Ethical compromise is part of participating in society, even the unemployed have to follow the law. Moral compromise is typically avoidable though, and some principled people will hold those morals equal to the value of a job.
... I'm sure I'll get flamed into oblivion for this but it's weird to me how the zeitgeist is anti-colonialism but also against enforcing borders and national sovereignty. I guess maybe they are okay with it unless you're a western nation. Whatever, there's no room for nuanced opinions anymore in modern online discourse.
I'll admit to being a weird outlier on this. I may not like what certain parts of the government are doing but I'd go work for the Voldemort companies in a heartbeat -- they just require in-office and aren't anywhere near me. I'd rather my nation develop the best technologies than let other nations do it.
Just look at how far behind the eurozone is by not making the right investments.
> I'm sure I'll get flamed into oblivion for this but it's weird to me how the zeitgeist is anti-colonialism but also against enforcing borders and national sovereignty
Let me try a reply that is perhaps a little more constructive than where this discussion went.
I think most folks would frame home rule and freedom of movement in the context of basic individual rights, not national rights. I suggest you might fare better explaining some context about why you believe they should care about nation states to begin with
"Freedom of movement" is a fairly loaded term, but is defined in the Universal Declaration of Human Rights is:
- "Everyone has the right to freedom of movement and residence within the borders of each state."
- "Everyone has the right to leave any country, including his own, and to return to his country."
Freedom of movement isn't the right to go live anywhere you want across any national jurisdiction (although some people/organizations would disagree, but that just doesn't survive any legal test). Every nation on the planet has very explicit rules around this. Even the unrecognized ones.
I guess I don't necessarily agree with elected officials systematically not enforcing laws that they don't agree with. It's hard to be rigid about this when there's things like Executive Orders, but I'm very pro "Change the law". And admittedly that's very difficult and it should be so -- new law should be hard to get to.
This process happens all the way from the local to national level, but we have terrible governance. Every structure seems to be abdicating its responsibility and overreaching in one way or another. It's precisely in these conditions that I think leaning into rule of law is more important than ever.
Otherwise we're just playing Whose Country Is It Anyway? Where the rules are made up and your rights don't matter. I also feel that this should be a compelling enough argument regardless of your political persuasion, but we've gone just about completely tribalist.
If it's any consolation, I'd have upvoted the original version of your comment before you appended the hand-wringing about karma points and personal opinions.
It's not lacking nuance to oppose both empire politics and nationalist jingoism. I don't think that the author is wrong for holding that opinion - I just think they shouldn't have joined Google if they sincerely believed it. Their personal conviction is not going to be enough to overturn Google's policy, even if they're the LeBron James of AI research. This was a pointless exercise in getting crushed under the boot that they claim to oppose. And anyone with the ability to add two and two together could see where this was headed.
Outliers like you are what the state counts on. Morally deaf, money and status-obsessed, easily placated with remote work and fat pension payments. As the middle class shrinks and the collective insecurity of a generation expands, more people like you will trade in their morals to make a "dignified" living doing horrible things. It speaks volumes to your character, but maybe you'd have a hard time hearing it from your boat with your hearing aids turned off. It's a tricky market for people that want to make genuine positive change in the world, and money talks.
If the author were to read these comments, I'd commiserate with them on the disappointment of the utopia left behind in the service of acquiring wealth and power. An abundant solarpunk future enabled and augmented by AI, hyped so well by early Google and Kurzweil, captured as capital to earn a buck for shareholders.
Would you consider that AI isn't a necessary precursor to an abundant solarpunk utopia? A lot of that hype was fictional, for better or for worse.
We're still struggling to proliferate the techno-utopia of digital computers, the internet, and cryptocurrency too. Each time this hype gets drummed up, the utopian stakes are raised to the nines, and the delivered technology does not bring us substantially closer to a more fair or utopian society. Pessimistically, all of these technologies were used to surveil, addict and intellectually degrade society, leading to the miserable status quo we all share in.
The technology is not the problem, it seems. I'll commiserate over the longing for utopia, but new technology has never been a panacea for structurally-embedded ills.
> The mechanism of individual technologies, both actual and possible ones, does not interest me much. I would not have to look into it if man’s creative activity were free, in a godlike manner, from being polluted by unknowledge—if, now or in the future, we could fulfill our goal in the purest way possible by being able to match the methodological precision of Genesis; if, in saying “let there be light,” we could obtain as a final product light itself, without any unwanted additives. However, the previously mentioned splitting of goals, or even the replacement of one goal with another, often an undesirable one, is a classic phenomenon.
I've migrating away for this and many other reasons.
I recall a slide from this year's i/o or next, where they put Palantir's name first on a slide of new partners. I was quite taken back by how loud and proud they are about it. Suppose I shouldn't be considering the CEO paid up the $1M to attend the inauguration.
It's nice to see someone willing to act on their principles. Unfortunately, as this person found out, most people are not that good.
More than that, I tend to think that so many Rubicons have already been crossed by Google and their ilk that it's pointless to think aiming at any narrow ethical target like this will make much a difference. So my appreciation for the author's ethics is tempered by a belief that, even if Google had done everything the author wanted on this one issue, the company would still be evil.
I still think there is almost no chance that companies so large, built on such a wide array of unethical activities, are redeemable. I don't see any real solution but wholesale dismantling of virtually all large companies in existence today, and putting in place far stricter regulations that prevent any companies ever becoming that large and powerful again.
I mean - the idea that any company employing non-trivial amounts of people "cares" in any sane sense of the word "care" is absurd.
Who is supposed to "care" about you? A single person, definitely not the CEO, is going to individually care about 20k+ workers, especially not 200k+ workers.
Even if you assumed an altruistic HR department, it's still going to be a faceless blob when you're a 100k+ company.
> I wanted AI ethics commitments to hold under pressure.
LLM chatbots, they way they are trained and have data collected for them today, are fundamentally unethical, regardless of whether Google sells services to the DHS.
No less importantly, Google's is a fundamentally unethical entity. After all, it says so right in its motto: "Do Evil".
Ok, that's not it's motto, it's just a pun on the old motto Google dropped when it became too starkly opposed to its practices, of mass surveillance and manipulation and censorship of search and content discovery results for commercial and political purposes, on its behalf and that of governments. This has been widely reported upon.
I will just give a few links regarding Google's complicity in the ongoing genocide in Gaza, as described by the UN special rapporteur for the Palestinian Occupied Territories:
> Jeff Dean ... and other senior employees pledged to “neither participate in nor support the development, manufacture, trade, or use of lethal autonomous weapons.” Google signed the classified deal, yet they remain.
> I think he could have stopped the deal, yet he did not. He remains, yet I think he should not.
Is Jeff Dean still considered a saint at Google? If so, how come this doesn't change that? The amicus was enough?
I sometimes think this way but I do wonder whether this would be true in a world where individuals demand more control over the impacts of their work. It seems that the current problem is that the labor of the 98% (or whatever number you like) is well-aligned with the demands of the 2% without principles who seem to have a strong propensity to end up in leadership positions. It reminds me of the simulations people run where in a high-trust world scammers do well, while in a low-trust world they die off only to then lead to a new high-trust world.
If enough people make decisions like OP and only perform work that they believe in (for one reason or another) perhaps that redirects the path to power for the 2% such that they need to act in the interests of the 98% rather than the other way around. I think OP was very late to the train here and I had made a similar career switch in the past for similar reasons but I'm happy to see other people making decisions that, if nothing else, will make them feel that they are in control of the impacts of their time and hard-work. Not sure how things will balance out in the end/distant future but we do our best and try to lead a life we feel good about living.
This is a long post about someone who has very obviously just gotten into politics. It is good for people to try and see how to impart change. Here are some constructive critical questions for the author:
1. Why no mention of No Tech For Apartheid or Google Workers United, who have been doing similar work for years?
2. What about all of the other police, DHS, and military contracts Google has been a part of? Did this problem really just start with the second (not even the first!) Trump presidency?
3. What does a focus on exclusively those at the top levels of a hierarchy, with minimal focus on incentive structures and wider systems, say about your theory of change? Was there a power analysis done, or was it assumed that "big title" = "powerful"?
Side Note: Incredibly insulting of James Dean to say email 3 CEOs.
I don't hold it against people for only caring about one or two issues. In addition to many plausible explanations, people like OP who take a stand are already doing so much.
I for sure agree about encouraging people to dig deeper and wider, and that absolutely none of this is new or is occurring in a vacuum, but (and I don't say this to single you out, this is just a long-held feeling I've had) I do also think it's important to encourage and commend anyone who reaches these conclusions, however late or nascent they might be. Especially someone with the backbone to speak out at and then quit a cushy, prestigious job.
Something something every journey, something something single step. For the author (and for all of us, really) I hope it's one of many. And I think they should be proud of this particular step :)
The main answer would probably be: because he worked for Google Deep Mind as opposed to Google in general, so the killerbot and mass profiling concerns were immediately relevant to him and his work.
Your comment seems to be whataboutism. The article was not an essay trying to prove that he knows the secret to being maximally effective at politics, or whatever it seems you were expecting?
Nothing but respect to TurnTrout for taking an action like this. The world needs more smart people who are willing to stand for what they feel is right, despite the pressures otherwise. Without that occurring more, our species is going to lose many impactful prisoner's dilemmas coming these next two decades.
This raises my respect for AI researchers a little bit too. I have often felt that the entire industry is pretty tainted to the core, and for better or worse that colors my opinion of the researchers.
Maybe I'm in the minority, but I thought it was gross to download pirated art for a student project when I was at Berkeley years ago. So it has been really sad to witness many of the most brilliant minds of this generation answering the siren song of disrespecting the collective effort of others to extract and resell residual value.
I'd guess TurnTrout doesn't agree on that framing, otherwise he probably would not have been at Deep Mind. But clearly he and I agree on other ethical positions; I am nothing but glad to see him stick to his principles here.
Thank you for your praise.
> I'd guess TurnTrout doesn't agree on that framing, otherwise he probably would not have been at Deep Mind. But clearly he and I agree on other ethical positions; I am nothing but glad to see him stick to his principles here.
FWIW I agree that creators should be compensated (but evidently it wasn't a deal-breaker for me joining). I think it's bad how little that has happened.
I joined GDM to work on AGI safety to reduce existential risk from AI. I consciously avoided work that would improve the raw capabilities of Gemini. When I joined, I made a trade-off and it's valid to disagree with my decision.
As your writing seems to allude to, life is an impossibly complicated series of trade-offs made with imperfect information, so I definitely wouldn't consider joining an AI frontier lab to be a priori bad. Even if I have some qualms with how the models are built.
If anything I was just trying to point out that, even with people we might disagree with, we deserve to show them recognition and respect when they make moves we do agree with.
But I doubt we disagree on much. Your writing is great, and it makes me sad to see this thread somehow disappear from the HN front page so fast. Either it triggered some internal flamewar detector--but without a flamewar--or someone didn't find it convenient.
I wish you the best of luck with your next endeavors, and look forward to seeing the value you add to this world.
I'll add some praise- thank you for taking a stand. I trust the employment will work itself out as it tends to for smart and principled people.
Apologies, but it would be good to add links for your anecdotes.
Not all of the readers of your comment have the appropriate context and know what you're talking about. I certainly don't.
Hmm, links for what? My student project? The decisions that face humanity that are pretty clearly modellable as prisoner's dilemmas?
Otherwise not sure what I could cite--I would assume most all on this forum know that AI is trained on the works of other people, without their permission to do so. I guess you could disagree with my framing, but I wouldn't think this requires a citation.
I think maybe my writing wasn't clear, and it sounded like I was referrencing some well known thing that happened at UC Berkeley. I have edited it to read more cleanly!
Yes; the student project you were talking about.
It wasn't clear to me it was _your_ project nor who pirated it. I was under the impression it was a well known scandal from your original, unedited comment.
Apologies if my comment sounded hostile; I was just asking for a clarification/more information on it.
No worries! Thanks for engaging. Fun to have the rare kindly-resolved internet discussion.
Props to the author. I left Microsoft due to their work with Israel to spy on Palestinians and record all of their phone conversations (in addition to other IDF collaborations). They ended up walking some of it back, but Satya was complicit.
https://www.theguardian.com/world/2025/sep/25/microsoft-bloc...
So Microsoft will support Israel in killing Palestinians, but pull the accounts of International Criminal Court Prosecutors? Jesus
I salute you for your morality. If most people did this, those companies would change their position very quickly. After all, they only think about money.
Thank you but I'm no saint. I joined Microsoft knowing that they were incompatible with my values. I just told myself that I wouldn't work on those types of projects and try to stick to my values. I told myself that I could just be "professional" and keep my head down.
But, as I learned more I was filled with a deep sadness and shame. It was tearing me apart. Every day I got up to go to work and was getting sick to my stomach.
Now I just feel lost. I spent my life building a career in this industry and it feels like I have nowhere else to turn. Microsoft is not the only company actively complicit with horrible crimes and human rights violations. I've loved technology since as far back as I can remember, but looking back with the knowledge I have now, I just feel dirty. I honestly don't even know what to do at this point, I do have a job now but when I look at who funded us, it's every bit as bad as working at Microsoft. I feel like I need to walk away from it all and start a food truck or something.
It's more rare, and definitely less lucrative, but there are opportunities in software engineering for non-profits, open-source, and smaller non-tech companies. If you've made your blood money and don't need top n% pay, it's something worth checking out.
As you astutely point out, it's not easy, many companies are complicit, or at least the people and groups funding them.
Have you looked at the healthcare or green energy domains with smaller companies/start ups?
Respect!
I haven't finished reading, but I should note for the group that the author doesn't represent the Anthropic situation accurately. Emil Michael, the undersecretary of war/defense for research and engineering went on the All-In podcast (1) and explained, for quite some time, exactly what happened. You should go listen to it (2). In essence, in the negotiations, Dario kept coming back saying "well, if you need it, call us, we can redline things as needed". This happened over and over and over. Emil's point was that a major conflict, if it were to occur, might happen on the 30 minute clock of an ICBM, and all due respect to Dario, the national security apparatus just doesn't have an allowance in that 30 minutes to call him for a redline. Emil felt Dario had demonstrated he wanted ultimate control via line item veto, and was willing to trade up to and including the survival of the nation for that veto. And a government cannot be expected to pay for that sort of behavior from an entity domiciled in their jurisdiction.
Now, Dario is going to win a decent slice of the economic pie. But as an military acquisitions matter, I gotta say, I have to agree with the undersecretary's position here, and yeah, it makes sense to document a company's undesirable behavior, and in certain circumstances push that information to others. Not the first time the government has found a company under contract acting in a way that appears to be counterproductive to the government's obligations; there's a whole database full of this stuff (3).
1) Without a doubt, All In is a friendly crowd for Emil, but I think that actually made it easier for him to get more nuanced facts out in this case because he wasn't spending a lot of time defending malignant attacks.
2) This link should jump to the relevant segment https://www.youtube.com/watch?v=gzwRflcLPAA&t=2479
3) http://www.ppirs.gov/
That doesn't address the supply chain risk designation? Or the ban on all work with claude for contractors, incl. non-government work?
Even if they disagreed, this was clearly retaliatory for not capitulating. The government could have just decided to go with a different provider.
I cannot think of a less trustworthy source of information than Emil Michael on the All-In podcast. Why on Earth would you take his statements at face value? I grant that nobody in this situation is totally trustworthy, but if it's his word against Dario's I'm believing Dario 10 times out of 10, and it's astonishing and dismaying that anyone might do otherwise.
Keep in mind that Emil Michael's job at Uber was to slander and destroy their political enemies. What makes you think he hasn't continued exactly as before?
Wait. Do you trust what he said?
Principled people have become so rare
Only when the bank account reaches 7 - 8 figures.
Even at zero dollars and a billion dollars they feel rare. Honestly I am surprised how many tech people still work for Facebooks and Googles of the world when they actively have used their platforms against most of their values. These are skilled people that can still make a living elsewhere.
I Think it’s because when you have enough money to never have to work again and you can move to some place nice, you can isolate yourself and your family from the rest of the world, like most rich people do.
So that’s what, 4-5 years at a society destroying tech co in exchange for lifelong freedom away from the people you have actively made life worse or impossible for.
It seems many people can live with that and in fact will jump at a chance to do it.
No need to downplay it. There are many for whom it's never enough.
And there are others who left years ago and warned us. Back-patting isn't all or nothing, and I think it's reasonable to praise those who left when they saw the writing on the wall more than those who leave only once they see their company materially supporting state killings.
Fair.
I did not know about TurnTrout till now and thanks to HN community now I know. I am beyond impressed as to how he lives by his principles. I don't think I have this kind of courgae or confidence and thankfully I don't work for Google or have to quit it. On the other hand we have smart people who have choose to work for companies like Palantir and are proud of it.
It is not about courage or confidence, it is more a calling, where few are called and that often turns out to be just one person, or so it seems, since everyone finds their excuse to not take a stand against the oppressor and for the oppressed.
The author will have experienced every logical fallacy and TurnTrout has done well to document his efforts, rather than just the crimes of his adversaries.
A calling of conscience is hard to articulate initially, it is conscientious objection, to the war machine and the system of finance that necessitates the empire that needs the war machine.
The author starts off with a high degree of authority in that he actually worked for Google Deepmind, however, nobody will listen to him, so he has to seek higher authority to carry the truth forward. But he could have gone to the pope, anyone short of Jesus Christ (freshly teleported back) and it will be a no, everywhere.
Logic and reason does not help when insanity and money have taken over, which tends to happen during wartime.
The conscientious objector may not believe in god, however, they will consider themselves answerable, 'vain' enough to care about their legacy. There is that desire not to be found out, doing the 'devil's work', generally in workplaces where everyone is glad to get on with the 'devil's work'.
Why do so few people make a stand? Why didn't hordes of five-eyes people walk out the day after Snowden did his good deeds?
It varies by individual, however, the author is vegan, which means he has already 'dared to be Daniel, dared to stand alone', albeit only in a lunch queue in a meat eating world. This requires living according to principles, and serves as target practice for war-time conscientious objection.
Also important is a certain level of independence, the guy with a mortgage and a couple of kids, underwater on the car, with maxed out credit cards cannot conscientiously object. Nor can the guy counting down the last few years to retirement, and then the very young lack the articulation.
Only a few have the 'warchest' to embark on an open-ended single-person campaign to confront power with the truth. We owe a lot to these people, particularly the 'failed whistleblowers' that don't make it to 'whistleblower' status, because the media then makes the story about their situation, not what they conscientiously objected to. Props and respect to the author for documenting his journey and taking the first step.
Note the 'first step' is classic 'hero's journey', where the call is initially rejected, but then a journey into the extraordinary world is made, where the challenge is to bring back what is good from the extraordinary world to the ordinary world, for the benefit of all.
Respect for sticking to your beliefs. Just out of curiosity though why do people not want smart AI weapons? I would much rather have an onboard AI that can discriminate between unarmed civilians and military assets, seems irresponsible to not... Is a dumb sea mine that blows up everything somehow better than a smart sea mine that knows to not blow up sometimes?
"You're absolutely right, that wasn't a military target—it was actually a girls school. It won't happen again!"
I realize that's not a great argument and was definitely tongue-in-cheek, but given there's still a lot of debate about the accuracy of AI for far more mundane tasks, my personal perspective is that until we have LLMs and such that are truly, demonstrably far more accurate than humans, with true reasoning and judgement capabilities, they don't belong where lives are at stake.
I wouldn't want an LLM-underpinned machine running anesthesia during a surgery; why would I want an LLM-underpinned military apparatus that is deciding the lives of far more? I wouldn't, not in their current state.
In a hypothetical future where we truly trust incredibly smart AIs or LLMs or whatever "smart" technology it is for driving weaponry, okay - if it's truly necessary; I abhor war and the death and destruction wrought by it.
In my mind, though - even if we get to that future where there's some vastly superior technology to the LLMs we have today, which can judge and reason, then I'll have a bunch of other questions, like understanding the motivations of said technology, because I suspect it'll be something much closer to AGI, and that opens a whole separate can of philosophical worms.
Not only that but an AI could follow any orders it is given (whether it hallucinates that's a different story) whereas humans may oppose. The friction of orders having to go through a chain of command through its completion is a feature not a bug and that adds responsibility. AI can take no responsibility by itself and with 'moving fast break things smart weapons' that responsibility can be further shoved under the rug.
> "You're absolutely right, that wasn't a military target—it was actually a girls school. It won't happen again!"
What if it was the target though? AI may be more capable than we're giving it credit for (especially the AI accessible to the US and other governments). The attack on the girls' school coincided with Purim and I don't believe it was a mistake. I think it was the opening salvo by a radically religious Zionists (Christian and Jewish).
I don't think the biggest problem with AI weapons is that they make "mistakes", I think the biggest problem is that they allow people who want to kill civilians the ability to accurately do just that.
So you would prefer weapons that cannot reason at all?
> "You're absolutely right, that wasn't a military target—it was actually a girls school. It won't happen again!"
Most likely this event happened due to a bad targeting system that wasn't smart enough, if it had a better llm underpinning it (assuming it had any) maybe those lives would have been saved. More reason for more smart people like the author to work on these systems.
The bottom line is a human should have to fire the gun and answer for it
> So you would prefer weapons that cannot reason at all?
Um, yes? It's bad enough humans are murdering each other. At least a human can in theory be held accountable for pulling the trigger. The last thing we need is an unaccountable ralph loop reasoning about which schools and churches to bomb every time it wakes up.
> if it had a better llm underpinning it
Ah yes, the "LLMs are intelligent, you're just not using the newest fanciest model" fallacy. This time with innocent lives on the line. If only we used ChatGPT-8.9 instead of 8.8, those poor kids would still be alive today.
President Trump reveals the US Space Force has CAMERAS that can read the NAME TAG "Mohammad" trying to reach any nuclear dust in Iran
"We say: 'Mohammad something is there with shovels.'"
"We have cameras that can read the badge of the person. 'Mohammed Something.'"
"WE'RE WATCHING. If anybody goes there, THEY GET BLOWN UP."
"Eventually, we'll take it."
They have also admitted they saw the children’s flower chalk drawings too. And they double tapped.
Mainly I don't want them because I don't want a machine instantly making life or death judgements. It might be fine if it mostly informs a human who ultimately pulls the trigger, but having that failsafe is important in keeping machines from going on killing sprees.
I also want people to be held accountable when they do unjustified killings. AI weapons make it FAR too easy to simply pass off a killing as a "woopsie doodle." It's just not acceptable to say "The algorithm made a mistake, version 23 will do better".
I don't have a problem with the AI providing additional information to it's user, but when that's incorporated into a weapon it's a short distance from that to completely automating the killing.
That's why I'm completely against AI weapons.
Responsibility will always have to lie with the user, but if the weapons are more advanced and safer then why is that bad? I like that autonomous vehicles like Waymo are 10x safer then human drivers, even if a 'Machine' is making decisions.
Because they aren't "safer" for the folks getting killed.
Historically, the folks doing the murdering just don't care about the folks they kill, so "safer for the killers" isn't a win for most of us.
"More precision" isn't about killing less folks, it's about making it safer and easier for folks who kill folks to do that work.
So those of us who dislike killing don't like these tools because we consider them to be immoral.
That is a great point and one I never considered. The folks doing the murdering shouldn't have their jobs be easier. They aren't safer for the person getting killed.
> Responsibility will always have to lie with the user
Being fully autonomous makes it hard to identify exactly who that user is and is easy to dilute responsibility. Perhaps someone was added to a kill list by mistake. Maybe some internet hi-jinks tricked an AI into falsely identifying someone as a kill target. Perhaps it's the case that someone was in the 5% of a 95% confidence of identification. I'm not a fan of putting killing into the hands of something known to get things wrong 1 in 20 (95%), or 1 in 100 (99%), or 1 in 1000 (99.9%).
> but if the weapons are more advanced and safer then why is that bad?
It's yet to be proven that they are "safer" as they become more advanced. There's also a question of "safer to who". It's technically "safer" for a soldier to shoot first and ask questions later, it's obviously not safer for the villagers.
False identification, which is an absolute part of AI, doesn't make these weapons safer for anyone.
> I like that autonomous vehicles like Waymo are 10x safer then human drivers, even if a 'Machine' is making decisions.
Waymo has the reverse bias. If anyone dies as a result of waymo it's gone horribly wrong.
AI weapons are designed exactly to kill, if they don't kill when they should something has gone wrong.
A good thought experiment for this is to replace guards at jewellery stores or banks with a robot. Which one would you feel more safe? Do you trust that the robot won't randomly shoot you if you act funny on the street?
Jevons paradox, but for violence.
Right now there's a lot of killing that's only limited by the risk of collateral damage.
Flipped around, the best deterrent against major wars is mutually assured destruction.
> Just out of curiosity though why do people not want smart AI weapons?
Because the minute you defer decision making to a machine you eliminate human accountability.
Any time the robot is used to do something illegal or immoral, you can just say "it was a bug" or "that's not what I really meant". It enables plausible deniability.
We see this every time AI is allowed to make decisions. That robot promoted violent extremism in your feed? It misidentified a black man? It disproportionately chose men in hiring decisions? shrug Its the algorithm. Who can argue with the algorithm?
Likewise any attempt to codify liability is typically avoided at all costs because the purpose of those systems is what they do.
> I would much rather have an onboard AI that can discriminate between unarmed civilians and military assets,
Has this been proven to actually occur? There have been lots of reports that similar systems in Israel that use AI, struck hundreds of targets that were civilian. They did not discriminate.
https://en.wikipedia.org/wiki/AI-assisted_targeting_in_the_G...
It’s true that humans are unreliable, even if we could guarantee a consistent set of moral and ethical standards as well as the appropriate tools and authority to enforce them at all times. We do, however, have to deal with 100% of the implications, one way or another. Machines don’t. Vasily Arkhipov and Stanislav Petrov had to consider the world they would condemn their children to if they pushed the button, the missiles didn’t. These are imperfect devices made by imperfect beings, whose rules and goals can be changed more or less at will by whomever wields them. Offloading the responsibility of pulling a trigger or trampling someone’s legal or human rights to such a machine while insulating those in command from consequences threatens to make things worse at a speed and scale I don’t think we’ve seen before in all of our dark history. Meanwhile, it further concentrates immense and unchecked power in the hands of a relative scant handful of elites, all of whom exhibit the lack of ethical and moral fiber we were talking about to begin with.
Teaching something or someone to pull a trigger is trivial compared to teaching them not to. I have no reason to believe that automated weapons and surveillance will be more reliable than our world leaders or that the world will be safer with them in it.
> Is a dumb sea mine that blows up everything somehow better than a smart sea mine that knows to not blow up sometimes?
Yes. If smart mines lead to belligerents laying more of them because they're "safer."
A regime can only be as oppressive as the people following its orders. There are only so many protestors you can order your army to kill until your troops realize they are killing their own friends and family, at that point they either desert or turn their guns against you. This puts a check on what even the most evil tyrants can do. Not a perfect one, for sure, but still better than nothing.
Technology changes this by reducing the human involvement needed in killing. The Nazi gas chambers were a historical example — it turned out mowing down innocent civilians with machine guns was too mentally taxing for even devoted Nazis, so they invented a bloodless industrialized process where they can kill with greater efficiency while also sheltering the killer from the immediate effects of his actions. A win win for the Nazis, not so much for its victims.
Autonomous weapons are dangerous for this exact reason. They bypass individual moral decisions entirely. They are the perfect "just following orders" machine. They allow a small concentration of powerful people to impose force upon the populace without ever having to fear their own soldiers. It tips the balance of power massively in favor of those already powerful.
>Just out of curiosity though why do people not want smart AI weapons?
When the decision to kill another human being is made that should be in the hands of a directly accountable other human being, not an unaccountable machine developed in the basement of a private corporation.
And mines, both dumb and smart, in particular anti-personell mines are banned by the Ottawa treaty ratified by 162 countries. It's exactly the autonomous and fundamentally uncontrollable nature of mines, not just that they're dumb, that has produced countless of casualties long after wars were over. Can you tell me that millions of autonomous loitering munitions are not going to end up exactly like those mines still blowing legs off people decades after conflicts are over? And who is responsible then?
The mines got banned because they were dumb...
no they were banned because they're persistent. That's also the US position despite not being signee to the treaty. Making autonomous weapons intelligent only makes the issue worse because at least expertise was a limiting factor for most rogue actors. Do you know what will happen to millions of autonomous drones when the Russia Ukraine war draws down? Which hands they will end up in? You've created the capacity to kill people 2k kilometers away for a thousand bucks on a chip.
And when those inevitably backfire and start blowing innocent people up, are the CEOs going to prison, the politicians who ordered their production originally? No, nobody is going to claim they're responsible, and that's why it's insane to build a weapon that does not have a name, time, place and direct order associated with it.
Even the US military is already incapable of taking responsibility of blowing a school with 150 kids up. You want to hand these people systems that are unaccountable by definition, that are smarter than they are, when they already struggle to understand the consequences of their own decisions?
They aren’t training smart AI weapons to discriminate between civilians and soldiers. Even so, what happens when the soldiers are disguised as civilians? Or when the enemy forces civilians to serve or be a decoy?
Because it's going to be used in any possible way, and certainly not for the good of the people, as imply in your "naive" false dichotomy.
This is inspiring, thank you to the author, it must have a hard piece to write.
> It is also the story of something larger: how powerful people and institutions failed, one after another, to keep their AI ethics promises in the face of pressure.
I've worked at Google. These weren't promises, they were deceptions.
Google, like many companies, maintains the focus and motivation of their workforce via a stream of carefully orchestrated disinformation and half-truths. One of those half-truths is "We listen to our employees". They listen when it has no business consequence and they can tout themselves as being "good" or on the side of the workers. They don't whenever the politics or economics of the situation presents even the slightest resistance.
Kudos to the author for fighting for change. But the dark patterns in Google run too deep for them to be uprooted by a polite exchange of white papers and emails.
This is no longer on the front page of HN despite 200+ points and having been posted 1 hour ago.
There’s a flame war detector, this has certainly triggered it and pushed the thread down the rankings.
I don't see any evidence of a flame war in this discussion, in fact there is an overwhelming consensus in the comments on the side of the author.
You really believe that?
I saw it on the front page.
This is where I first saw it too, but 10 minutes later it was no longer there.
They move uncomfortable subjects off the front page pretty quickly these days.
Can’t have the sheep getting too rowdy.
We know now that Google is a sycophantic company whose DEI initiatives were all fake and dropped once Trump got elected.
This somewhat naive initiative was bound to fail. The good news is that the AI military products won't work, except perhaps for blowing up a girls' school.
Here are CEOs falling over themselves to support Hassabis' regulatory capture proposal:
https://xcancel.com/sundarpichai/status/2077086951833063580#...
https://xcancel.com/satyanadella/status/2077063479232795024#...
https://xcancel.com/elonmusk/status/2077415601610297535#m
It is an exclusive club and we are not part of it.
Come on, ALL companies are psychopaths in their behavior.
Their only objective is to make money to shareholders. So anything that gets in their way is fair game. They are relentless optimizers.
After all, most people holding GOOG, MSFT, META, or others don't care how I happens, they just want their 401k fund going up. It's like that by design.
Saying AI military products wont work is the dumbest thing i've heard in awhile
Search for Anduril drone failures in Ukraine and several Lattice software failures. And perhaps capitalize "I".
How is Maven working in Iran?
Search for Ukraine mid-range strike capabilities in the last 1-2 months. It's using homegrown Ai systems to great effect, re: Crimea is cutoff to the point gas is heavily rationed or not sold to civilians at all.
Thats like saying cars wont work because they get flat tires
It's almost like Anduril has little to no background in building military soft/hardware [0][1]. The modern day charlatans are alive and living very well.
[0] https://www.techtimes.com/articles/320165/20260711/army-ar-g... [1] https://www.techbuzz.ai/articles/anduril-s-autonomous-weapon...
Bro, they're not going to work. I promise you lol
Well done, great effort! Unfortunately, most people are weak in their comfort.
It is honorable but ultimately hasn’t democracy spoken wrt the issues mentioned in the post?
For better or worse millions of Americans voted for the guy doing the deportations.
I also find it difficult to reconcile not using AI for weapons. If war is inevitable AI presumably would at least ensure you are on target.
Agreed, I feel like the people who say no to AI weapons haven't actually presented a real argument (that I have heard) besides terminator bad
Regardless of whether AI is involved, many people simply don't want to work on weapons systems, or on non-weapons technologies that are intended to become part of a weapons system.
Thats fine, but a serious luxury belief IMO
Not really, it is fairly common belief. And American war system is not in the pro-democracy, defense only or whatever else business anyway. It is in pro-warcrime, pro-stupid-wars, pro-dictators business.
I think there's some nuance here that a lot of these convos dont make plain - are we talking about just using 'ai' in the military broadly, or are we specifically talking about fully autonomous weapons systems?
In the former case, I would agree with you completely, I havent heard any arguments beyond 'I dont like working on military stuff'. But if we're talking fully autonomous weapons, that's a different story. And further muddying the waters is the fact that the former is obviously a step along the path to the latter.
AI helps people scale. If being "on target" means that you find people when they're eating dinner so you can kill their entire family, then using AI to scale that up is a terrible thing. That's exactly what Israel did with Lavender. They literally had a kill list called "Where's daddy?".
https://news.ycombinator.com/item?id=39918245
As opposed to what though? Again, I assert violence is inevitable.
AI is the end of the funnel and I actually think the mass surveillance at the beginning of the funnel is a bigger issue. The concentration of power is the problem. If you or I could defend ourselves against state level actors with AI weapons, then they would be a good thing (imho). AI and mass surveillance at the state level create a power imbalance that is a major threat to human rights and should be resisted as much as possible.
When folks who live around me vote for things I find unethical, then I don't feel bound by those decisions.
Also, your position on weapons development is premised on the idea that at lease some of the folks developing the weapons are on your side and always will be.
That might not be true at all.
Same goes for Putin's policies of course.
> For better or worse millions of Americans voted for the guy doing the deportations.
Uh, the idea of democracy is not that voting reveals universal preferences or something. It is ok to disagree with whoever gets elected, and continue working towards an alternative you believe to be better. In fact, democracy depends on that; otherwise, why not have a single election with no term limits? There is supposed to be ongoing difference of opinion.
> I also find it difficult to reconcile not using AI for weapons.
You can't reduce "AI for autonomous lethal weapons" to "AI for weapons".
It does appear to be anti-democratic given everyone knew Trump’s platform when he was voted in. However the great unifying fact of politics is that everyone only really believes in democracy when they win.
This idea that once a party gain a little more votes, people who oppose it should cease to exist and cease push for own values ... somehow applies only to the right.
When center or left wins, somehow, magically, same logic dont apply.
What a weird strawman
Literally you "It is honorable but ultimately hasn’t democracy spoken wrt the issues mentioned in the post?"
Ethics commitments brushed away under pressure are just Marketing.
> Plus, a pledge is only worth the credibility behind it. When someone signs “I will not support the development of lethal autonomous weapons,” then stays while their company sells unrestricted AI to a military that wants exactly that, they teach every counterparty a lesson: these safety people will not act, even at their own brightest line. The next commitment they make is worth less. Eventually it’s worth nothing.
This is why I haven’t taken the internally employed (and visibly public) AI safety people seriously for years: they always back down, because they value that fat paycheck and fiscal rewards and associated status over their proclaimed ethical values, red lines, and boundaries. Their actions betray their every word.
That’s also why I push that we remember these names, the ones who fought back as well as those who lied through their teeth. We should remember everyone willing to harm others to protect themselves, and hold them accountable for their misdeeds. In an ideal world, people like Jeff and Sundar and Stuart would be blackballed so hard they’d never be able to get so much as a help desk job ever again, nevermind any position of leadership. It’s why I proclaim them to be monsters, giddily sacrificing the anonymous other for personal vainglory and the avoidance of standing on any sort of principle.
The OP did everything right and demonstrably proved the entire apparatus is rotten to its very core, to the point of infesting and poisoning the very entities professing to stop its worst excesses (IASEAI). Why we give these ghouls the slightest bit of credibility anymore is beyond me.
> Even if Google had adopted your Framework, the Pentagon would have refused
> I agree. xAI would still have given over their AI. But if Google had given signs of independence earlier, it could perhaps have built a coalition with OpenAI and Anthropic.
This is like saying that Google and Apple might build a coalition to prevent App Store regulation. These are competitors, they all see moral flexibility as an advantage. They're not going to take a moralist stance if their federal protection is predicated on federal cooperation. All of these FAANG businesses have already bent the knee in anticipation of this, xAI is just riding the coattails of the federal quid-pro-quo.
When you go to work at a megacorp, you're always leaving your ethics at the door. Yes, there's an attractive pie-in-the-sky fantasy that Apple does care about human rights, or that Google isn't evil, but they're always just lies. I disagree with a lot of GCP's customers, but I'm still shocked that a DeepMind employee would make it this far in the career pipeline before seeing the soylent green get made.
Competitors often wish they didn't have to compete.
For example, a cartel is "a group of independent market participants who collaborate with each other and avoid competition in order to improve profits and dominate the market"
It's something that could happen in this world, through the collective action of employees, that corporate strategy can be changed.
I hope he becomes the seed crystal around which crystallization of peaceful use of AI happens.
> through the collective action of employees, that corporate strategy can be changed.
I just don't agree with this. I'd like to, but the corporate strategy is not being set by rank-and-file engineers. Sam Altman, Dario Amodei, Elon Musk - these are the executives that are steering the ship. You can change the culture internally and cause a lot of strife, but typically you don't have control over the strategy. If the executive says you're giving API access to the NSA, then that's what is going to happen.
apple and google are regularly on the same side against eu regulations
Working anywhere requires ethical compromise… a 2 person company can disagree on ethical questions
Ethical compromise is part of participating in society, even the unemployed have to follow the law. Moral compromise is typically avoidable though, and some principled people will hold those morals equal to the value of a job.
On the upside, I guess maybe DeepMind is hiring.
... I'm sure I'll get flamed into oblivion for this but it's weird to me how the zeitgeist is anti-colonialism but also against enforcing borders and national sovereignty. I guess maybe they are okay with it unless you're a western nation. Whatever, there's no room for nuanced opinions anymore in modern online discourse.
I'll admit to being a weird outlier on this. I may not like what certain parts of the government are doing but I'd go work for the Voldemort companies in a heartbeat -- they just require in-office and aren't anywhere near me. I'd rather my nation develop the best technologies than let other nations do it.
Just look at how far behind the eurozone is by not making the right investments.
> I'm sure I'll get flamed into oblivion for this but it's weird to me how the zeitgeist is anti-colonialism but also against enforcing borders and national sovereignty
Let me try a reply that is perhaps a little more constructive than where this discussion went.
I think most folks would frame home rule and freedom of movement in the context of basic individual rights, not national rights. I suggest you might fare better explaining some context about why you believe they should care about nation states to begin with
Fair comment and thank you.
"Freedom of movement" is a fairly loaded term, but is defined in the Universal Declaration of Human Rights is: - "Everyone has the right to freedom of movement and residence within the borders of each state." - "Everyone has the right to leave any country, including his own, and to return to his country."
Freedom of movement isn't the right to go live anywhere you want across any national jurisdiction (although some people/organizations would disagree, but that just doesn't survive any legal test). Every nation on the planet has very explicit rules around this. Even the unrecognized ones.
I guess I don't necessarily agree with elected officials systematically not enforcing laws that they don't agree with. It's hard to be rigid about this when there's things like Executive Orders, but I'm very pro "Change the law". And admittedly that's very difficult and it should be so -- new law should be hard to get to.
This process happens all the way from the local to national level, but we have terrible governance. Every structure seems to be abdicating its responsibility and overreaching in one way or another. It's precisely in these conditions that I think leaning into rule of law is more important than ever.
Otherwise we're just playing Whose Country Is It Anyway? Where the rules are made up and your rights don't matter. I also feel that this should be a compelling enough argument regardless of your political persuasion, but we've gone just about completely tribalist.
If it's any consolation, I'd have upvoted the original version of your comment before you appended the hand-wringing about karma points and personal opinions.
It's not lacking nuance to oppose both empire politics and nationalist jingoism. I don't think that the author is wrong for holding that opinion - I just think they shouldn't have joined Google if they sincerely believed it. Their personal conviction is not going to be enough to overturn Google's policy, even if they're the LeBron James of AI research. This was a pointless exercise in getting crushed under the boot that they claim to oppose. And anyone with the ability to add two and two together could see where this was headed.
Outliers like you are what the state counts on. Morally deaf, money and status-obsessed, easily placated with remote work and fat pension payments. As the middle class shrinks and the collective insecurity of a generation expands, more people like you will trade in their morals to make a "dignified" living doing horrible things. It speaks volumes to your character, but maybe you'd have a hard time hearing it from your boat with your hearing aids turned off. It's a tricky market for people that want to make genuine positive change in the world, and money talks.
If the author were to read these comments, I'd commiserate with them on the disappointment of the utopia left behind in the service of acquiring wealth and power. An abundant solarpunk future enabled and augmented by AI, hyped so well by early Google and Kurzweil, captured as capital to earn a buck for shareholders.
Mad props for pushing as far as they could push.
Would you consider that AI isn't a necessary precursor to an abundant solarpunk utopia? A lot of that hype was fictional, for better or for worse.
We're still struggling to proliferate the techno-utopia of digital computers, the internet, and cryptocurrency too. Each time this hype gets drummed up, the utopian stakes are raised to the nines, and the delivered technology does not bring us substantially closer to a more fair or utopian society. Pessimistically, all of these technologies were used to surveil, addict and intellectually degrade society, leading to the miserable status quo we all share in.
The technology is not the problem, it seems. I'll commiserate over the longing for utopia, but new technology has never been a panacea for structurally-embedded ills.
> The mechanism of individual technologies, both actual and possible ones, does not interest me much. I would not have to look into it if man’s creative activity were free, in a godlike manner, from being polluted by unknowledge—if, now or in the future, we could fulfill our goal in the purest way possible by being able to match the methodological precision of Genesis; if, in saying “let there be light,” we could obtain as a final product light itself, without any unwanted additives. However, the previously mentioned splitting of goals, or even the replacement of one goal with another, often an undesirable one, is a classic phenomenon.
- Summa Technologiae
Oh, absolutely. But it could be used right and well in a broader solarpunk framework :D
Google and Palantir are quite proud of their new found friendship
https://cloud.google.com/solutions/palantir
https://www.palantir.com/partnerships/google-cloud/
I've migrating away for this and many other reasons.
I recall a slide from this year's i/o or next, where they put Palantir's name first on a slide of new partners. I was quite taken back by how loud and proud they are about it. Suppose I shouldn't be considering the CEO paid up the $1M to attend the inauguration.
It's nice to see someone willing to act on their principles. Unfortunately, as this person found out, most people are not that good.
More than that, I tend to think that so many Rubicons have already been crossed by Google and their ilk that it's pointless to think aiming at any narrow ethical target like this will make much a difference. So my appreciation for the author's ethics is tempered by a belief that, even if Google had done everything the author wanted on this one issue, the company would still be evil.
I still think there is almost no chance that companies so large, built on such a wide array of unethical activities, are redeemable. I don't see any real solution but wholesale dismantling of virtually all large companies in existence today, and putting in place far stricter regulations that prevent any companies ever becoming that large and powerful again.
Powerful
I think a man who makes that kind of choice is seriously admirable. I could never have made that choice myself.
am I the only one that finds it very hard to read this font?
If you do not like what Google is doing, just leave as I said before [0]. They don't care about you.
[0] https://news.ycombinator.com/item?id=47931336
I mean - the idea that any company employing non-trivial amounts of people "cares" in any sane sense of the word "care" is absurd.
Who is supposed to "care" about you? A single person, definitely not the CEO, is going to individually care about 20k+ workers, especially not 200k+ workers.
Even if you assumed an altruistic HR department, it's still going to be a faceless blob when you're a 100k+ company.
> I wanted AI ethics commitments to hold under pressure.
LLM chatbots, they way they are trained and have data collected for them today, are fundamentally unethical, regardless of whether Google sells services to the DHS.
No less importantly, Google's is a fundamentally unethical entity. After all, it says so right in its motto: "Do Evil".
Ok, that's not it's motto, it's just a pun on the old motto Google dropped when it became too starkly opposed to its practices, of mass surveillance and manipulation and censorship of search and content discovery results for commercial and political purposes, on its behalf and that of governments. This has been widely reported upon.
I will just give a few links regarding Google's complicity in the ongoing genocide in Gaza, as described by the UN special rapporteur for the Palestinian Occupied Territories:
https://novaramedia.com/2025/07/02/tech-giants-and-british-b...
(or get the report itself: https://www.ohchr.org/sites/default/files/documents/hrbodies...)
> Jeff Dean ... and other senior employees pledged to “neither participate in nor support the development, manufacture, trade, or use of lethal autonomous weapons.” Google signed the classified deal, yet they remain.
> I think he could have stopped the deal, yet he did not. He remains, yet I think he should not.
Is Jeff Dean still considered a saint at Google? If so, how come this doesn't change that? The amicus was enough?
tldr: scientist discovers reality
If half of us were half as principled as this guy, the world would be a much better place.
No, they would just walked all over by the 2% who have no principles whatsoever. Kind of like what's happening already.
Principles are only good up until they enable you to be systematically victimized.
I sometimes think this way but I do wonder whether this would be true in a world where individuals demand more control over the impacts of their work. It seems that the current problem is that the labor of the 98% (or whatever number you like) is well-aligned with the demands of the 2% without principles who seem to have a strong propensity to end up in leadership positions. It reminds me of the simulations people run where in a high-trust world scammers do well, while in a low-trust world they die off only to then lead to a new high-trust world.
If enough people make decisions like OP and only perform work that they believe in (for one reason or another) perhaps that redirects the path to power for the 2% such that they need to act in the interests of the 98% rather than the other way around. I think OP was very late to the train here and I had made a similar career switch in the past for similar reasons but I'm happy to see other people making decisions that, if nothing else, will make them feel that they are in control of the impacts of their time and hard-work. Not sure how things will balance out in the end/distant future but we do our best and try to lead a life we feel good about living.
Yeah I don't even judge them poorly for making that decision, I just come to different conclusions.
Now I think there are plenty of other reasons not to work for Google or Facebook so I guess I end up in the same place for different motivations.
Don't play fair in a game others cheat.
tldr: scientist tell us that defending one's values needs courage and strong will.
This is a long post about someone who has very obviously just gotten into politics. It is good for people to try and see how to impart change. Here are some constructive critical questions for the author:
1. Why no mention of No Tech For Apartheid or Google Workers United, who have been doing similar work for years?
2. What about all of the other police, DHS, and military contracts Google has been a part of? Did this problem really just start with the second (not even the first!) Trump presidency?
3. What does a focus on exclusively those at the top levels of a hierarchy, with minimal focus on incentive structures and wider systems, say about your theory of change? Was there a power analysis done, or was it assumed that "big title" = "powerful"?
Side Note: Incredibly insulting of James Dean to say email 3 CEOs.
I don't hold it against people for only caring about one or two issues. In addition to many plausible explanations, people like OP who take a stand are already doing so much.
I for sure agree about encouraging people to dig deeper and wider, and that absolutely none of this is new or is occurring in a vacuum, but (and I don't say this to single you out, this is just a long-held feeling I've had) I do also think it's important to encourage and commend anyone who reaches these conclusions, however late or nascent they might be. Especially someone with the backbone to speak out at and then quit a cushy, prestigious job.
Something something every journey, something something single step. For the author (and for all of us, really) I hope it's one of many. And I think they should be proud of this particular step :)
The main answer would probably be: because he worked for Google Deep Mind as opposed to Google in general, so the killerbot and mass profiling concerns were immediately relevant to him and his work.
Your comment seems to be whataboutism. The article was not an essay trying to prove that he knows the secret to being maximally effective at politics, or whatever it seems you were expecting?