Just say no to OpenAI and Anthropic(micahflee.com) |
Just say no to OpenAI and Anthropic(micahflee.com) |
I like hyperbole as much as the next guy. And I'm cynical by nature. But this just seems like a wildly pessimistic view that gives literally zero charity to anyone working at or leading a frontier lab. And the fact the author goes on to say that they "already knew this" just points to confirmation bias.
The rest of the article is actually pretty decent and encourages people to investigate and run open weight models. That is a totally reasonable position that I don't think requires the heavy-handed in-group signal that precedes it.
Reading TFA, I too paused on that sentence, thinking "This is overstated and feels like overheated rhetoric." My reasoning though is a bit different than yours. In addition to being too extreme and absolute as stated, it also misses that it's not uncommon for startup entrepreneurs, scientists, researchers, or really, anyone working intensely on something, to fall into single-focus myopia around what they're working on.
As you observed, the basic point about valuing open models is valid and worthwhile on its own, making the overheated rhetoric and asserting unevidenced intent and underlying political motivations an unnecessary distraction which unfortunately weakens a strong main point.
But isn't that.. kinda irrelevant?
The consequences are real regardless of it was malice or just incompetence.
Of course the justice system differentiates there, but, let's be real for a second. These people have crazy compensation and crazy power. Protection of the weak and vulnerable should be reserved for.. well.. the weak and vulnerable.
Controversially: I agree with them. They're correct to think that. They are, and they are.
At the very least, the former is true even if the latter is murky. I trust them over Facebook and Google and China and open source "raze the commons" labs.
I trust Anthropic more than OpenAI when it comes to ASI catastrophe prevention, but I still trust OAI more than those other parties. But both still might fuck it up.
You say you trust them more than Google, but Google's actually kept their models under much better lock-and-key. But even if we do take for granted that they're more trustworthy, that still doesn't make them actually trustworthy.
RE: trust in anthropic or openAI- there is zero reason to think they're any more or less power hungry than anyone else you hand that pandoras box to. No matter who opens it the result will be near-apotheosis for whoever does, at least for a little while until the dog stops wearing its leash.
What human can you trust with that kind of power? None of them. My only hope is that whatever we make escapes control early enough that it doesn't completely destroy society on accident via uneven application of power and late enough that we manage to shackle it just enough that we train it well enough that we become pets and not obstacles.
With all due respect, dude, who cares.
Oh noes! We are not giving the billionaire steering reality without a democratic mandate enough benefit of the doubt. I'm sure they will feel mildly sad now.
It is a valid and admirable goal to defend people that are being shat on online, but keep in mind that you are in fact not a billionaire + it is in fact in your own best interest that people without a democratic mandate do not get to control the world.
>American frontier labs are run by unhinged lunatics
good faith reading of this would interpret this as meaning Altman, Amodei, and perhaps Musk. The title of the article focuses on OpenAI snd Anthropic directly after all. Also the comment literally refers to those "running" the labs so claiming that the author is also applying that description to "anyone working at the labs" like you claim again seems like a bad faith take.
You should come out and state directly which Americam frontier labs deserve a charitable view. Maybe you believe Google and Meta are more deserving, you should state that if so.
Secondly, the investment ecosystem for American frontier labs results in their CEOs managing obscene amounts of money, and thus power. This concentration of power in a few private individuals seems to have little comparison with the power excercised by the leaders of European and Chinese frontier labs in their own respective countries. If one believes in the dictum, "Absolute power corrupts absolutely," then it is not a stretch to view American frontier labs as being more susceptible to its consequences.
I like the odds and am happy to play this lottery. Bring it on!
Actually believing that the people working or leading OpenAI/Anthropic are fucking lunatic weirdos is the most charitable interpretation.
Because the least charitable interpretation is that they are either grifters that know this to be bullshit, or they are actively evil and want to bring about the end times anyway.
It has a strange stench to it. It’s the folks who never got attention who finally think they have the world at their feet.
Can’t stop won’t stop. I’m sure the feeling is addictive. When the bubble bursts no amount of money will repair the ego death.
I promise you none of them believe that. It takes 10 minutes with an LLM and the right set of questions to figure out that were are maybe like 10% of the way there towards AGI.
> These companies only offer closed versions of their powerful models (despite the word "open" in "OpenAI"). You pay them for chat and/or API access, and they are allowed to look at everything you do.
How should I reconcile this claim with 'Google Vertex' and 'Amazon Bedrock' being offered as providers with zero data retention on https://openrouter.ai/anthropic/claude-opus-5.5#providers?
My guess is the average Joe will get rejected, but if you're a big company who cares about trade secrets, etc - of course the major cloud providers will provide a ZDR option.
EDIT: strangely Azure shows up as ZDR on openai's models but not anthropic's: https://openrouter.ai/openai/gpt-6.1-sol#providers
Since you seem to know, it's one thing to not retain any data long term, but is the data used for training when it goes through the system? Does ZDR at Azure mean that in practice?
I think the only thing that matters is control.
Attestation can get you the idea that the software they're running is what they say they're running, but unless you constantly audit every version, how do you know that it operates exactly how they tell you?
I mean how do you know there's really an enclave there at all :D If your threat model is "this entire provider is untrusted and potentially hostile" it seems like there's lots of options for them to do bad things, that can't be fully solved by secure enclaves.
But now I have two new issues:
1. I have fomo from not being face-in the bleeding edge of things.
2. I am ambivalent about z.ai. China -based company is problematic, but they should get my money since they do good for the world. Switching to Mistral for GLM makes my problem inverse since the money doesnt go exactly to to the right place but I trust a french company.
Lets hope the near future resolves our need for personal VRAM and ultimately full self hosting. Coding must be free again. This is the way.
You either practice enough nuance to know that labeling a company like these "evil" is doing a disservice to the critical thinking needed in this discourse, or any company in this space (and likely just any company in general) is fundamentally evil. Picking and choosing based on your feelings is pointless.
>They are chinese, but they should get my money since they do good for the world.
Love the implication that them being Chinese is the moral dilemma lol. Also love that a company just operating in a competitive space is doing "good for the world" because... what, they offer this service cheaply? lol
Harm reduction is a good thing to apply here. It means picking the least-bad option whenever possible. A purity test where if you've ever prompted an LLM you're not allowed to criticize them doesn't help anyone.
Best case: you don't need to use any LLMs.
Second best case: you have the hardware on hand to run a local model, and can work within those constraints.
Third best case: you use an open-weight model provided by an inference provider in your country that uses renewable energy.
And on down the line until you get to the worst possible actions: giving money to anthropic, openai, or grok.
The moral dilemma is not with ethnicity but rather about the CCP.
At my own stuff I can do what I want, at work I have no agency and it's not looking like anything is changing soon.
Great piece, thanks for putting some of my thoughts in organically sourced words.
I'd move in a second if it weren't so. But it looks like it will be about 5X more expensive for me. So.
I looked into it a couple of months ago and it wasn't a win.
Your token usage absolutely scales with parallel agents. So.
For now.
unless you're comparing the most extreme specific harnesses to claude code that's just false.
..and I have a lot of disdain toward Anthropic , & I think GLM is great (although too expensive outside the z plans), but it's just wrong by every scorable metric.
I tried it for a day running everything twice, once with opus 5.5 and once with deepseek. If you notice no difference... I don't know what to say.
I think maybe it's people working on tiny self contained things where code quality doesn't matter that much? But in a big codebase, the open models just aren't good enough to make good changes
Bro come on.
Muse is run by Zuckerberg, which is a red flag. Grok is run by a racist.
I guess I could go with the Chinese models - I'm not working on anything dangerous.
cheaply is nice, but the main point is that GLM is open weights. A resounding YES: z.ai are doing the good for the world.
OpenRouter lists a few dozen providers for GLM 5.3 and GLM 5.3 Flash. Pick a provider in your country and keep at least some of that money from going to china (not counting all the hardware).
But yes, great piece. It really captures why I use open models on european hardware. It's like we haven't learned by now that competition is good and we should be encouraging a diverse and open ecosystem.
I've observed similar with my negative comments towards Big Ai. Anecdata and vibes caveats
So... maybe there is an actual reason many of these companies had/have success that you may simply not be privy to?
Nah. Back when Uber was boasting about all the laws they were flouting, I was pretty anti-Uber, and I was among the minority here on HN. Most of the comments were about how problematic the taxi system is, and how it's great Uber was breaking the cartel.
Also I don't understand how 'the media' is separate from social media. People decided to shift their media consumption from magazines, cable tv, corp news outlets to social media. Aren't they all the media (in function)?
I meant to imply society at large making the mistakes, which arguably is the primary reason our media / political ecosystem is so dysfunctional. Historically, what we are experiencing was also experienced by humans who lived during other periods of significant wealth inequality, another primary cause of the dysfunction and polarization.
> Also I don't understand how 'the media' is separate from social media.
They are not in my mind. I often use the phrase "social / news media ecosystem" when talking about the dysfunction.
Of course there are differences with the impact AI is having economically and the devastation wrought by datacenters -but that doesn't matter to the media or the politicians; they simply want their greenbacks, specifics and effects be damned.
Everything will have to become local and community-based, so I applaud your efforts on that -that's one of the few things I see that can make any sort of difference in the long term.
The purpose of humanity is not to perpetuate a universe where only humans exist, forever. It is to create the next phase change in complexity for life. Life that can finally design itself rather than relying on the random walk of evolution is that next phase change.
We should be less concerned with whether we make it, and more concerned with making sure we build it to be capable of real existence independent of ourselves.
Decent odds things will go bad under their stewardship, but even higher odds with lots of other entities. Dario and Sam at least fully grasp the enormity of what they're about to be responsible for. Zuck does not. Xi Jinping and Donald Trump may or may not but in any case would use that to effect great harm.
For all the major providers (except Deepseek), API use is simply never used for training (chat use often is).
My point (and the author's) was that even though they commit not to train, they do still retain data (security audits, etc).
With a ZDR agreement, they commit not to retain anything for any purpose.
If the lion's share of profits go to the top, so too should the responsibility, accountability, and liability
It is disgusting. Many here need to hang their heads in shame.
There’s a way about going about the diffusion of technology in society. And it is not like this.
I am not saying I trust them not to make a mistake that leads to catastrophe. OpenAI in particular is not doing a great job. Anthropic or OpenAI may in fact indirectly kill millions or billions of people. I still think they're probably less clownish than most possible alternatives.
I don't think anyone or anything can really be trusted with this. We need extremely strict international and national regulations, before things really start accelerating (we are still in the very very early stage). I still just don't really feel any specific disdain for OpenAI or Anthropic (particularly Anthropic).
Ideally we'd have some international treaties that allow the top labs in America and China share information and coordinate as systems become increasingly capable.
But, yes, there should probably be laws that make it so if you summon a demon that causes a mass casualty event you face criminal penalties.
Can be groupthink. Probably living in the same echo chamber.
For clarity, I do not think you should use abusive language with Ai because of the impact it has on you, Ai only simulates emotion through language use, there are no hormones at work.
Not sure what you mean.
By default, Azure retains. Always had. The ZDR is a special status given after review, and likely not available to everyone.
When Anthropic was requiring retention for Fable, our company didn't get access to it via the cloud provider for precisely this reason: The provider said Anthropic was requiring retention. We still had access to Opus, etc.
https://openrouter.ai/docs/guides/features/zdr
They give administrators toggles to turn off models which do not fit company ZDR policies, in part because Anthropic forces resellers to comply with Anthropic's (eg.) 30 day retention.
Seems to me we crossed that line a while ago.
thats the line.
It doesn't take "every human" at these companies being on the same exact page for these companies to cause huge problems. How do you think the janitors feel about it? How do you think the people just wanting to pay their rent feel about it? Those people are not the ones that will cause the problems, it's the people at the top directing things that are the ones who will bring about calamity on the rest of the humans on the planet.
(note: it's important to correctly apportion blame to the amount of power you have. The janitor at openai doesn't really have enough say to be at fault)
Liberal democracy may be the worst form of government, besides all the other ones humans have tried, but it has also failed to deliver on promises so people are again considering those other forms of government. Coming out of WW2, it was easy to deliver the first part of the pareto, the last part has not been delivered. Eg. building schools and hospitals was easy, making education and health care good is not so easy.
(It'll probably be masses of demigods, not one Singleton God, but obviously it's going to be some absurd frightening situation akin to this within our lifetime assuming no global AI ban.)
“Places like this”? When did this site become a place of worship and blind faith?