Why I Left Google DeepMind
turntrout.com
turntrout.com
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.
Not all of the readers of your comment have the appropriate context and know what you're talking about. I certainly don't.
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.
> 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.
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.
>The stereotypical activist action is to make a petition. But Google had already ignored a large petition on this issue. Plus, Google’s executives likely hardened their company against stereotypical organizing tactics. Sit-ins, strikes, even a mass of Google engineers quitting: I deemed all of them ineffective (if I could even pull them off).
https://www.theguardian.com/world/2025/sep/25/microsoft-bloc...
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.
Have you looked at the healthcare or green energy domains with smaller companies/start ups?
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.
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
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 if they disagreed, this was clearly retaliatory for not capitulating. The government could have just decided to go with a different provider.
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.
> 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.
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 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.
... 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.
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.
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.
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.
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.
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.
> "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.
"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.
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.
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?
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?
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.
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.
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.
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.
Yes. If smart mines lead to belligerents laying more of them because they're "safer."
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...
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.
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.
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.
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.
Can’t have the sheep getting too rowdy.
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.
How is Maven working in Iran?
[0] https://www.techtimes.com/articles/320165/20260711/army-ar-g... [1] https://www.techbuzz.ai/articles/anduril-s-autonomous-weapon...
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.
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.
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.
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".
When center or left wins, somehow, magically, same logic dont apply.
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.
Mad props for pushing as far as they could push.
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
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.
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...)
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.
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 :)
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?
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 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?
Principles are only good up until they enable you to be systematically victimized.
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.
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.