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.
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.
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.
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
Even if they disagreed, this was clearly retaliatory for not capitulating. The government could have just decided to go with a different provider.
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?
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.
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.
> "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.
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.
"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.
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.
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.
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.
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?
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.
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...
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.
> 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.
I chose to be vulnerable and honest knowing what kind of responses I would get and the guidelines of this board call for assuming good faith. It's discourse like yours that is inflaming and ruining our society.
I didn't even endorse any specific actions here, I'm just not meeting your moral threshold based on your interpretation of some pretty thinly-detailed comments that I made.
You also can't influence the game if you're not a player so opting out of working in the entire defense industry is probably against your interests.
> I'd go work for the Voldemort companies in a heartbeat […]
being honest and vulnerable about being a shitty person does not beget or earn you any kind of mutual respect. we are simply not the same.
And I know what it's like to live in a place without adequate law enforcement and/or military.
Personal insults and calling me mentally ill. Are you projecting?
Apparently willingness to work at Palantir or Anduril or other defense contractors makes me irredeemably evil. I'm sure you're real fun at parties.
Tbh, I'm glad though. I'm really great at what I do and have had a long and successful career. We need more hiring filters and I don't want coworkers with deeply skewed perspectives, inadequate objectivity and who cannot coexist with those of different world-views. Self-filtering is the cheapest and best kind.
and I don’t want coworkers that are okay with killing children in Gaza or conducting mass surveillance and calling it a “different world-view”.
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
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...)
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.
People almost universally agree that stripping citizenship from naturalized citizens is way too far (no cite here, just recalling a recent read).
[0] https://www.pewresearch.org/short-reads/2024/11/22/most-amer...
On the other hand, people dying is a real concept. Their deaths because some people think that other people shouldn't be around and appealing to national sovereignty to enforce that is a leap into the absurd.
At the end of the day, societal institutions exist because we choose by action and by decision to support them. If they aren't supported, the world and the trust network become dramatically reduced in size.
EDIT: to the dead comment below -- your comment doesn't have to meet my approval -- the comment is simply at odds with reality and is a thought terminating cliche.
Feel free to take it in a more general framing: if you don't want people you don't support, while they are in power, killing people, take time to understand why others might be upset. Not only is it your civic duty, its also part of being in society.
TFA's point is really getting at that -- who is accountable for scaling bad actions and actors?
It's a stretch that any software Google/DeepMind/etc is selling to DHS is allowing / helping them to scale the murder part of their operations.
In fact, usually software translates to "less boots on the ground" which one could then assume would decrease the number of encounters like those highlighted in the article.
1. US aid abroad to help stem the conditions that drive many of the migrants to flee their homes. Build up institutions, enforcement, and anti-corruption frameworks so these nations can better build themselves. I'd even be open to hawkish approaches -- frankly it's been surprisingly to see the positive responses to Venezuelan interference by the US.
2. Immigration enforcement needs use-of-force accountability and less lethal approaches. The specific clashes between protesters and ICE come from actions that appear practically designed to encourage confrontation. Agents that kill innocent people should also be held accountable, meaning civil liability reform is necessary.
3. Ensuring we have a shared reality -- hold organizations accountable when they lie or disinform. Example - the majority of immigrants aren't actually criminals, nor are they "stealing jobs" in any coherent way -- studies show that many immigration policies have been net positive on communities. So when Fox News or similar organizations bring up random anti-immigrant commentary, dressed as news instead of entertainment, there really should be consequences for misinforming.
4. Targeted investigations instead of street sweeps. While Kavanaugh stops may be legal, like many enforcement policies over the years they are less effective unless the goal is "enforce stereotypes broadly."
> Good was in her car, stopped sideways in the street, which led Ross to circle her vehicle on foot. Other agents approached, and one ordered her to get out of the car while reaching through her open window. Good briefly reversed, then began moving forward and to the right, into the direction of traffic. At this point, Ross was standing several feet away at the front-left of the vehicle which was turning away, when he fired three shots, killing her.
[0] https://www.mprnews.org/story/2026/01/08/questions-follow-af...
[1] https://en.wikipedia.org/wiki/Killing_of_Ren%C3%A9e_Good
No one disputes this. She and the others were deliberately antagonizing agents first. They were not random innocent third parties. She deliberately ignored the lawful orders of armed officers and drove towards them. In almost any other situation involving police, no one would be surprised if this happened.
> began moving forward and to the right, into the direction of traffic
You can watch the videos from neutral sources if you're concerned. There was no one in her way.
With essentially any country, and even vaguely similar circumstances
https://www.theguardian.com/world/2021/nov/05/belgian-police...
https://www.theguardian.com/world/2020/aug/20/belgian-police...
And this is just a random example country (with excellent beer), you will find worse than this in most/all European countries, especially in the last 10 years or so.
On top of that, they are acting aggressively and violently in broad daylight solely to terrorize immigrant communities. The chilling effect is very visible in hospital systems right now - I have seen far less Hispanics in the hospital for medical emergencies, and that includes people who are in this country legally.
Wake the fuck up and stop trying to whatabout your way to a more comfortable mental state.
typical police activity protected by impunity in this country.
It sounds like it is normal for majority of political elite as well as voters, since this is not top of political agenda in this country.
It is not normal from my point of view, but I am not sure what can I do about this.
conjuring empathy in others is not easy, but it is worth it.
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?
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.
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.
That is a big presumption.
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.