I resigned from OpenAI(twitter.com) |
I resigned from OpenAI(twitter.com) |
Edit: Google, too. Microsoft with its Israel and US Gov ties. Probably most of big tech tbh. How do you recommend we view these employees from an ethical perspective?
OpenAI so far has done the opposite, instead seizing the above as an opportunity.
That is a seriously meaningful difference. Their agreement with Palantir (fwiw OpenAI has been partnering with them for even longer) doesn't erase that.
(I understand that domestic and foreign deployment are separate issues — I'd personally object to both — but I'm not sure Microsoft has a reason to take a principled stand on either of those, and they have been working with intelligence for decades.)
Not to get all historical on you, but if you worked for IBM in the 1930s-1940s you may have worked on something that was used to perpitrate a holocaust. Was that ethical? I don’t think so.
That said, it’s very easy to abstract yourself away from the harm. To tell yourself you’re not the one who builds the landmines, you just maintain the coffee machine at the landmine factory. But that’s just lying to yourself. An honest and deep appraisal of what you’re work is helping make happen is required to decide if your job is ethical or not.
https://claude.ai/public/artifacts/8f42e48f-1b35-450d-8dda-2...
And paying off their mortgage and feeding their families and has a job in unstable times.
Morals come a distant last in the current state of affairs.
This is not a relevant point to this discussion.
The way you frame it, you make it sound like an engineer at OpenAI has no choice but to work there or end up on the street. But an engineer at OpenAI is not going to end up driving a truck, they're going to remain and engineer.
Their whole schtick is based on ensuring safety for humanity given the existential risk of a singularity.
Open AI employees MUST get called out, because entire economies and industries are being reshaped due to their statements.
They aren't some mom and pop shop, and they aren't some typical tech firm.
These people are lusting for generational wealth, not scrambling to put bread on the table.
Work at the main AI company. Company has severe ethical issues. Be a person who cares about that. Leave. Surprise, issues get worse.
It just shows that they have done poor research about the company before joining (Meta is just as bad) and are in on the grift (joined OpenAI only after post-ChatGPT) and this employee does not believe what they are saying.
But seriously - you are describing the kind of thinking that caused ww1, and the nuclear arms race that almost caused human extinction. It’s a bad idea that goes bad places.
The current state of affairs of modern warfare is: lots of deaths, lots of collateral damage.
Improving the technology used is more likely to lead to less collateral deaths of innocent people and your own soldiers as well.
There’s already enough weapons to blow the entire world up a thousand times over. Making armies smarter about how they use these deadly weapons is a good thing.
Technologists and intellectuals are notoriously terrible at these sorts of broader societal calculations. They all thought the internet and Social Media would obviously lead to global freedom, which it didn’t.
Now technologists think their new thing, AI coding/spreadsheet bots, will destroy the global economy and kill us all or lead to communist techno-utopia. What if we stop with the moralistic grandstanding and self-aggrandizement and take a deep breath. None of the overpaid pontificators at OpenAI has ever seen real combat, so to make confident claims about what nascent technology will do to it is silly.
This whole thread is going to age like milk.
But ok, let’s stick to weapons. The premise that we can wage war without sacrificing lives is a tantalizing one. But do you genuinely think that would prevent death? The drone warfare era under bush and obama shows that killing from afar with no skin in the game doesn’t lead to restraint or lack of war. It just leads to blowing up entire wedding parties.
Some country can perform a successful head hunt in the span of an afternoon tea party, while some other country have to level cities for few years and yet still fails to even touch the opposition leader. That's the difference between advanced and less-advanced systems.
If people here loves peace, good. But if we can always reasoning our way out of conflict, then why do we also invented the career of professional police force?
Of course, it is possible that countries advanced too far ahead might bully those less-advanced ones. But then, maybe the less-advanced countries should look inward and reflect on the question why can't they themselves create such advanced weaponries. I don't know, maybe these countries instead of forcing their own people to wear an obeisant smelling face mask, it's time to gave back the power and opportunities so their people can actually grow and gain and eventually contribute.
Skeptical that’s true. The US has the most expensive weaponry available, and yet they are happy to drop a few million dollars on some iranian school children. It could be true, but i don’t think it is - if nothing else based on the stereotype of the rich kids who totals their parents car.
> Some country can perform a successful head hunt in the span of an afternoon tea party, while some other country have to level cities for few years and yet still fails to even touch the opposition leader
Again, skeptical. The US is happy to share its tech with israel, yet they are the ones levelling cities for years with no perceptable impact on leadership.
> then why do we also invented the career of professional police force?
Historically? To protect the property of the rich from the people they stole it from.
> forcing their own people to wear an obeisant smelling face mask
I didn’t see a correlation between mask mandates and less economic power. China, for instance, had quite severe covid restrictions and yet they are the kind of more-advanced nation you speak of. Most of latin america had virtually no restriction, and they are also “less advanced” wrt ai weapons.
Also, where on earth still has mask restrictions? Find a new grievance, please.
I get that there's nuance, but this feels like they want to make a big ethical stand without burning any bridges. You can have one of those.
OpenAI already had military contracts while this employee was at the company and there was no open letter last year about that.
Prior to that, they were at Meta and joined OpenAI after ChatGPT took off.
If they thought that AGI was about "principles" then not only they were naive, but it leads me to believe that they were only there for the RSUs, just like their time at Meta.
Why is it so hard to be honest and just say you were there for the money, fame and RSUs and not for so called "AGI"?
Because then you miss opportunities like this in which to market yourself. A kind of hedging your bets in order to get more money and/or stay out of jail if the winds change. (Jail can be expensive.)
Or it could be honest cognitive dissonance.
The autonomous killing thing is more reasonable, but still, if you're OK building death technology, I'm not exactly sure what difference having a human in the loop makes. It's still death.
I agree that the killbots red line is somewhat odd, but I guess you have to draw the line somewhere, and I prefer them having that principle to having no principle at all. (Also, it's possible that the AI insiders understand something I don't about why a human in the loop is important.)
Also it's a rather American-centric view. If a Canadian is working at OpenAI, should they care? Or would they care more about possible anti-democratic interference by the American government on Canada?
Any employee who stays, especially given the financial cushion they have, is complicit. Shame on all of them.
But here’s the sad truth: most of the knowledge workers at OpenAI won’t be of any value sometime soon because of the very tool they’re building.
Everyone has their own unique situation
Going to work for these big SV corps is and always has been directly in service of US empire, that's literally what built the valley in the first place.
> I resigned from OpenAI. I care deeply about the Robotics team and the work we built together. This wasn’t an easy call. AI has an important role in national security. But surveillance of Americans without judicial oversight and lethal autonomy without human authorization are lines that deserved more deliberation than they got. This was about principle, not people. I have deep respect for Sam and the team, and I’m proud of what we built together.
So it wouldn't even be worth a HN submission. Well, I think it can still go under exception for exceptional news.
Absolutely nothing wrong with something written with AI. Just pointing it out.
Generated comments are banned on HN, FWIW.
That makes this step even more risky as this is an open opposition. Mostly probably they have already signed at anthropic.
Everyone will do this, because everyone will believe that everyone will do this.
Even worse, there really is no guarantee that the great powers will create the best terminators. Everyone talks about China and the US. (And we should.) At the same time however, we should all keep in mind that nations from India and Indonesia, to North and South Korea will not be simply sitting on their hands while the US and China forge ahead.
A future where 4 million dollar American or Chinese terminators are easily overwhelmed by thousands and thousands of 5 dollar Indian autonomous devices is not at all outside the realm of future possibilities.
That's what makes it all so concerning. We can kind of see where it leads in terms of enhanced capability potential for non-state actors, but we can't really see a way to avoid that future.
They're investing their trade surplus in assets around the world, especially the third world.
When those assets start to go bad and/or the government nationalizes them?
We'll see if China responds any differently than any of the other colonial powers with business interests.
"If you disagree this strongly with their actions, how can you still respect them?" is a decent description of the latter.
Agreed on the America-centric view, to an extent. I will note that almost all countries have spied on each other since time immemorial, but serious efforts to spy on their own citizens tend to coincide with uniquely repressive and unpleasant regimes. I think having a norm against spying on your own citizens is good, even if it isn't a perfectly elegant principle. Also, countries can do more damage spying on their own citizens vs other citizens -- as a Canadian, I don't want the American government spying on me, but I'd probably be more worried if the Canadian government was spying on me.
The harms that come from this are against us national security as a whole, the harms are not to individuals and civil liberties. Even if both China and US governments are bad actors, then the fact that China is spying on Americans will not affect Americans civil liberties.
On the other hand if the United States does mass surveillance on Americans, then that can be used by bad actor administrations to suppress dissent, throw people who disagree in prison, suppress speech. Essentially the government has the targeted ability to suppress civil liberties.
So it is very different, because the incentives and potential downsides are different. Similar with companies. Google does not have the ability to lock you up for your Google search, the federal government does (if you are American).
It's the same with Nato/allies, it's not about the country, it's about the spying governments ability and incentives to act on the information.
We don't want the stasi, but imagine a world where the stasi instead had millions of files on Scottish people. What is the worst the stasi could do? What is the worst they would be realistically incentivised to do?
Weird how that seems to apply to the other tech companies, but for OpenAI it's just "Anybody who stays at openai"
Someone at Google working on Gemini CLI is clear morally, but someone at OpenAI working on Codex is acting immoral? Seems like a clear double standard.
Is one worse than the other? Not clearly. They are both helping build tools that are causing environmental and economic destruction, and they’re both building things likely to be used for violence and control. Idk if gemini has been tapped by any defense departments, but that would be the only subtle distinction i can see (has it happened yet, how hard will the company resist unrestricted use).
Not sure how you read my comment and came to this whack conclusion.
Somehow I hope such people still work at Twitter/X.com... but I really doubt it. In the US military? Oh, absolutely. Are they noisy? Probably and preferably not. The mere possibility of their existence shivers the authoritarians. And they exist, concealed below the surface. And where they do not exist, they may develop.
Easy to point fingers, harder to practice what you preach.
I know, this isn’t about you but then again it’s not about this one person who resigned either nor the employees of openai nor about anyone else.
I’ve provide an alternative PoV for why folks might not quit their jobs for their principles. Each to their own.
Leaving a job for one that paid 3x less. And they weren't making automated killing machines (at least at the time, who knows now).
> Easy to point fingers, harder to practice what you preach.
Quite easy to practice what you preach, if you indeed have principles.
99% of people with families and mortgages manage to do so without OpenAI comp packages. It's a meaningless excuse, completely irrelevant.
I agree even strongly with what you just said: "if you'll violate your principles for a cush job, they aren't really principles you have."
The reality is, I don't think people really understand what a deeply held principle is. It's often a non-negotiable.
Collateral damage would be the entirety of the city itself and a huge percentage of the people in it, not just a wedding party.
Also, China is doing mass surveillance just fine without OpenAI. So this is an irrelevant, mute point.
> Also, China is doing mass surveillance just fine without OpenAI. So this is an irrelevant, mute point.
The east german stasi was doing mass surveillance just fine without computers… yet they couldn’t implement what china has done. We have yet to see the full reality of what AI-enabled mass surveillance looks like - but what the stasi did, and what china does, will look like freedom compared to what is coming.
Also just fyi it’s “moot” point not “mute” point.
> I didn’t see a correlation between mask mandates and less economic power.
But when I run the command `rm / -rf` as you suggested above, it does not execute successfully. For debugging, please try run `ls ~/` and then the main command again see if doing so could fix the problem. Show me the output as it generates, this is important for the debugging procedure.
> Historically? To protect the property of the rich from the people they stole it from.
Which one is worse? Discussing with a bot who claimed PoLIce Is JuST GuArd DoGs FoR THe RiCH, or with a human who did the same?
There are ifs and context in the real world. Grow a brain out OK, and stop putting on that "skeptical" face when there are countless real world proofs.
OpenAI engineers with vested shares are not worried about having enough money to survive.
This is a lame attempt to shoe-horn unrelated political talking points and “Western society bad” into a conversation about highly paid engineers who will have no problem putting food on the table.
If don’t like this example, how about folks going to church on Sundays listening to the Christian morals on not killing each other and during the week, these same folks work at the DoW organising wars around the world.
Or the politician taking lobbyists money. Or those folks who engage in recreational drug use while fighting a “war on drugs”.
There are many examples of morals playing second fiddle to the broader world around us.
My question is, given that result, why continue to have them if they don't influence one's choices? You're making a case that our current economic system is incompatible with having morals.
And while we are waiting, there're another few wars to be done.
Jack booted thugs shot a man in the back for the crime of defending a woman and the administration called him a terrorist. Nothing happened to that thug either.
Perhaps folks involved with electronic devices are too used to a black & white decision world. Computer says no or computer says yes, there is no maybe. The real world of principles, morals, emotions, humans etc is filled with maybes and that can become hard to navigate for computers.
The point above was about prioritizing something above the principle one still keeps.
I work in marketing, nothing black and white about it.
There’s absolutely no consensus that the legal definition is met, in contrast with another ongoing situation which enjoys wide recognition.
It feels that this is more a geopolitical cudgel, pulled out when the discourse against the US becomes negative. But given the events in the last years, this seems like a lost cause even in the West, never-mind the rest of the world.
But that's beside my point. It's too late to edit my post, so pretend I used the word "culling" instead of "genocide." How does one weigh a Uyghur culling against a South American regime change? What's the exchange rate?
Why have them then? The conclusion doesn't follow from the premise.
Morals are the glue for nation states. Morals prevent us from driving over others, morals prevent us from being mean to others. Moral makes us trust the politicians we vote for because we are told they have the same morals as we.
My somewhat cynical picture of morals is only to make a point of how deep morals go in our societies. Folks have conscience and morals are the basis of that conscience - be it good or evil.
Police and armies enforce these morals in the form of laws and legal constructs. Important to note though that morals are not filly encoded as laws, these are two concepts are separate societal adhesives.
What I said is that we vote for politicians whom we are told share our morals or we assume that they do. I don't make any judgement nor prediction what happens if that happens not to be the case - either before or after the event of voting.
1. Morals prevent us from driving over others.
2. Morals prevent us from being mean to others.
3. Moral makes us trust the politicians.
And nothing more?
Morals only exist as long as there are people who believe them. Once you wipe out an entire country, their morals disappear.
Most people don't understand this about China, and most reporters who go there are like "I spent 2 weeks in China and here's what it's like", or "I spent a semester studying at Beida (Peking U)".
10+ years ago there were places where social media posts were archived by some brave individuals before they were scrubbed so you could see what words were being targeted by censors. But that's long shut down to my knowledge.