Anthropic ran a weeks-long roadshow on how powerful Mythos is. They pointed to the danger, their controls, the capabilities, and practically begged the world to be scared of it.
Simultaneously, the current US regime realized there was a way to demand fealty from the AI labs. If they're so dangerous, don't we need to see them first? That will cost you, obviously. Standard extortion from the government, at this moment in time.
The labs get their marketing; the white house gets its pseudo-bribe. I hope nobody involved is confused about how we ended up here.
> I'd be more concerned with "your model can't be too woke" regulatory scenarios.
Honestly that's exactly where my mind went. We already see the current administration trying to censor free speech (e.g. Jimmy Kimmel, blocking/restricting press access to the White House unless you are pro-Trump).
I'm afraid of the potential to move in the direction of what we see in China where queries to LLMs referencing things like Tianenmen Square are censored (at best).
I'm not american but it seems like americans are MORE free to speak their minds now than before in terms of being banned/silenced by dominating online platforms
Not sure what you're trying to say here, but I get the sense you don't have enough information as to what's going on here (nothing wrong with that, I've been trying to tune out myself).
Regardless: "...an increasing number of travelers report being questioned about legally protected online speech when crossing the border."
> I'm afraid of the potential to move in the direction of what we see in China where queries to LLMs referencing things like Tianenmen Square are censored (at best).
As it was with "campus protests violate free speech!" from the folks who immediately turned around and banned voluntary diversity programs at universities.
As it was with "Twitter bans violate free speech" from the folks who bought it and banned @elonjet and the word cisgender.
Because it's incredibly frustrating to see the government remove women and men of colour from government websites, deleting climate data, and sending out violent mobs to round people up while people sit around saying their main worry is that regulatory bodies will move to make things "too woke." There's no "woke" equivalent to the insanity being acted out by the US administration.
why? Most regulations (ADA, affirmative action, etc.) fall into the "not woke enough" category of model regulation. Current administration aside, complaints of this sort are more likely. It’s absurd, really, to believe there would be a regulation governing a model being too woke; regulation itself is woke
Health codes and safety rules are woke, yes. I would have thought that as given. Debates over where you draw the line are absolutely a matter of wokeness.
The way freedom of expression is regulated today is generally woke. The WPFI is insanely woke.
"Woke" is just a dog whistle. It's used by anti-intellectuals on the right to signal their allegiance to whatever their dear leader says, and is used to say "I am triggered by this idea" regardless of what the idea is. Anything can be woke to these people, up to and including the 2nd Amendment.
it is actually perfect, you hear or read “woke” you can immediately turn around and know the level of intelligence in front of you requires immediate extraction from any further proceedings :)
it is the people that are buying the BS though… those wielding power will say whatever the F they want to rile up the people but end of the day it is the people buying this.
the outmost funny thing to me glancing at this from the sidelines (I stopped following news and am generally largely disconnected from “politics”) is that people that are buying the “woke” garbage are not realizing they voted for the most woke there is. entire current administration is DEI hires (this is literal, every single person working is DEI hire), cancel culture is raging, if they could they would cancel 50% of the population (this is especially funny to me, shows you how “small” people are), you got billions of dollars literally being stolen in broad daylight while bitching about nancy pelosi, you got two president’s sons pillaging the country while bitching about hunter’s laptop… so fascinating to watch america slowly disintegrate into what has become
> Most regulations (ADA, affirmative action, etc.) fall into the "not woke enough" category of model regulation
For sake of argument, let’s assume this is true. Those rules are still structured as laws, with boundaries and legal recourse. The precedent being set, that the President gets “voluntary” deference from private companies, is un-American and will be abused by the left.
Yeah, I saw several instances of important folks taking the Anthropic promotional campaign too seriously and this is what they got in return. I'd say internally people are cursing whoever's idea that was because clearly scaring people backfired.
I would wager they're cheering, because this builds the moat they don't otherwise have. Want to do business in America? Get government-approved. Can't afford the regulatory fees, or your government won't let you submit to foreign programs? Good luck!
Yes, this has been a steady play from the start. From the skynet fears, to the safety fears, now the it's to powerful fears. All of these have been a play to get the government to lock out any smaller or foreign competitors and build a moat where there otherwise would be none.
Moves like this make me wonder- What chance is there that these models are nationalized in the near future? What will happen to the investors/economy in such a scenario?
Nationalization often happens when growth ends. The Pennsylvania Railroad was private as long as the profits were rolling in. But once growth ended (because of cars and planes and buses and ....) the company went bankrupt. Then we ended up with Amtrak because the country needs a train system.
once it gets nationalized, it will be plagued from red tape. The model will likely look like how china is controlling their AI. It's not nationalized, but they have a complete tight leash on it
So nationalized models === more openly available and downloadable models? Seems the argument you're trying to make says "less leash" rather than a stricter one.
Routine corporatism and fascism is shameless to the point of being ho-hum these days. When the president has his own cryptocurrency and the federal government buys stock in this and that company for "strategic reasons", you're looking at a dystopia.
I suspect the accelerationists, who appear more fascist aligned, are the ones upset by this. They go so far as considering regulation a form of murder.
This is probably seen as a win for the Bostrom crowd and the more sane people in the middle. The issues to tackle are incompetence and corruption and that has little to do with AI.
This is a strong thread that's needs to be plucked on again and again and again.
Cory Doctorow had an excellent thread yesterday that touches on this:
> You could be forgiven for assuming that this is just about reining in Wall Street greed, but that it isn't an especially political maneuver. That's not true: antitrust is the most consequentially political regulation (with the possible exception of regulations on elections). Every fascist power defeated in WWII relied on the backing of their national monopolists to take, hold and wield power. That's why the Marshall Plan technocrats who rewrote the laws of Europe, South Korea and Japan made sure to copy over US antitrust law onto those statute-books.
The well moneyed interests are getting everything they want, for the faintest little bribe. For showing the obsequiousness, for showing fealty to the regime.
The monopolization of power, allowing markets to en taken over by worse and worse foes of democracy, needs to be stopped. Needs to have some limit. The post talks about how:
> Under the Correcting Lapsed Enforcement in Antitrust Norms for Mergers (CLEAN Mergers) Act, any company that was acquired in a deal worth $10b or more will have to break up with its merger partner if it turns out that these mergers were "politically influenced."
(OpenAI and Anthropic reached a similar agreement with the US in 2024, per the article)
Well, I guess people who wanted more oversight and regulation on models will be happy.
I'm not sure they envisioned models being interrogated to determine woke levels and their opinions on the 2020 election.
Not the person you are replying to: but I think that's the point.
I'll admit that one flew over me. Ironically, an LLM would've probably picked up on the sarcasm.
Color me unsurprised.
Anthropic ran a weeks-long roadshow on how powerful Mythos is. They pointed to the danger, their controls, the capabilities, and practically begged the world to be scared of it.
Simultaneously, the current US regime realized there was a way to demand fealty from the AI labs. If they're so dangerous, don't we need to see them first? That will cost you, obviously. Standard extortion from the government, at this moment in time.
The labs get their marketing; the white house gets its pseudo-bribe. I hope nobody involved is confused about how we ended up here.
What extortion are you claiming?
Are you claiming there will be a fee?
> What extortion are you claiming?
Universities: https://www.npr.org/2026/01/29/nx-s1-5559293/trump-settlemen...
Companies: https://news.bloomberglaw.com/esg/extortionary-intel-stake-s...
Law firms: https://www.lawfaremedia.org/article/the-law-firms--deals-wi...
Media: https://www.nytimes.com/2025/07/02/business/media/paramount-...
Why would AI companies be any different?
> Are you claiming there will be a fee?
I'd be more concerned with "your model can't be too woke" regulatory scenarios.
> I'd be more concerned with "your model can't be too woke" regulatory scenarios.
Honestly that's exactly where my mind went. We already see the current administration trying to censor free speech (e.g. Jimmy Kimmel, blocking/restricting press access to the White House unless you are pro-Trump).
I'm afraid of the potential to move in the direction of what we see in China where queries to LLMs referencing things like Tianenmen Square are censored (at best).
I'm not american but it seems like americans are MORE free to speak their minds now than before in terms of being banned/silenced by dominating online platforms
Not sure what you're trying to say here, but I get the sense you don't have enough information as to what's going on here (nothing wrong with that, I've been trying to tune out myself).
Regardless: "...an increasing number of travelers report being questioned about legally protected online speech when crossing the border."
https://www.nytimes.com/2025/12/31/travel/airport-border-pho...
The government literally removed research into trans-fats because of their transphobia. Much free speech
> I'm afraid of the potential to move in the direction of what we see in China where queries to LLMs referencing things like Tianenmen Square are censored (at best).
We are already there.
"Canva admits its AI tool removed 'Palestine' from designs: https://gizmodo.com/canva-admits-its-ai-tool-removed-palesti...
Or your model is not "woke enough"
I have very little patience for bothsidesism at this point.
Your side is not the only side.
Obviously.
"Bothsidesism" posits that the two sides are broadly similar. The last few years have debunked that concept pretty conclusively.
It also posits that there are only two sides.
In the US, that is functionally the case, and likely to remain that way.
Yes but OP made a good point (model censorship going too far in the name of "woke"ness) and you shut them down.
Yes, because it's disingenuous bullshit.
As it was with "campus protests violate free speech!" from the folks who immediately turned around and banned voluntary diversity programs at universities.
As it was with "Twitter bans violate free speech" from the folks who bought it and banned @elonjet and the word cisgender.
Are people not allowed to suggest models may be too censored? Is that idea censored?
Am I not allowed to suggest it's disingenuous bullshit to pull the "both sides" thing?
That's right. It was uncalled for. I see no evidence OP was making a bad faith argument, but you assumed that right away.
Because it's incredibly frustrating to see the government remove women and men of colour from government websites, deleting climate data, and sending out violent mobs to round people up while people sit around saying their main worry is that regulatory bodies will move to make things "too woke." There's no "woke" equivalent to the insanity being acted out by the US administration.
why? Most regulations (ADA, affirmative action, etc.) fall into the "not woke enough" category of model regulation. Current administration aside, complaints of this sort are more likely. It’s absurd, really, to believe there would be a regulation governing a model being too woke; regulation itself is woke
The ADA isn't about wokeness. It's about being able to live in society with a disability.
Limiting a business's ability to exist because they can't afford to accommodate a small percentage of the population is 100% about wokeness
Well, we've just now defined safety rules, health codes, paying taxes, and the like as "woke".
You'd be calling the First Amendment woke if we proposed it now.
Taxes? How?
I agree with the rest, sure.
Health codes and safety rules are woke, yes. I would have thought that as given. Debates over where you draw the line are absolutely a matter of wokeness.
The way freedom of expression is regulated today is generally woke. The WPFI is insanely woke.
> Taxes? How?
They, at times, "[limit] a business's ability to exist because they can't afford to accommodate a small percentage of the population".
> Health codes and safety rules are woke, yes.
I take it you never read the parable of the boy who cried wolf.
getting the vibe that english isn't your first language, and don't feel like arguing theough a language barrier
i don't begin to understand either of these points
fwiw i'm woke, happy to be woke, encourage wokeness
> getting the vibe that english isn't your first language
Well, that's a first for me.
Alternate theory: https://news.ycombinator.com/item?id=48023653
> Debates over where you draw the line are absolutely a matter of wokeness.
This is offensive in how trite it is
"Woke" is just a dog whistle. It's used by anti-intellectuals on the right to signal their allegiance to whatever their dear leader says, and is used to say "I am triggered by this idea" regardless of what the idea is. Anything can be woke to these people, up to and including the 2nd Amendment.
it is actually perfect, you hear or read “woke” you can immediately turn around and know the level of intelligence in front of you requires immediate extraction from any further proceedings :)
The terrifying thing is that we can't just ignore it anymore, though. These people are wielding the power of the world at a time of global crises.
it is the people that are buying the BS though… those wielding power will say whatever the F they want to rile up the people but end of the day it is the people buying this.
the outmost funny thing to me glancing at this from the sidelines (I stopped following news and am generally largely disconnected from “politics”) is that people that are buying the “woke” garbage are not realizing they voted for the most woke there is. entire current administration is DEI hires (this is literal, every single person working is DEI hire), cancel culture is raging, if they could they would cancel 50% of the population (this is especially funny to me, shows you how “small” people are), you got billions of dollars literally being stolen in broad daylight while bitching about nancy pelosi, you got two president’s sons pillaging the country while bitching about hunter’s laptop… so fascinating to watch america slowly disintegrate into what has become
ADA preceeded wokeness by at least 2 decades
> Most regulations (ADA, affirmative action, etc.) fall into the "not woke enough" category of model regulation
For sake of argument, let’s assume this is true. Those rules are still structured as laws, with boundaries and legal recourse. The precedent being set, that the President gets “voluntary” deference from private companies, is un-American and will be abused by the left.
I don't think I'm smart/intellectual enough to respond to this
or i just don't understand what you're saying
Yeah, I saw several instances of important folks taking the Anthropic promotional campaign too seriously and this is what they got in return. I'd say internally people are cursing whoever's idea that was because clearly scaring people backfired.
I would wager they're cheering, because this builds the moat they don't otherwise have. Want to do business in America? Get government-approved. Can't afford the regulatory fees, or your government won't let you submit to foreign programs? Good luck!
Yes, this has been a steady play from the start. From the skynet fears, to the safety fears, now the it's to powerful fears. All of these have been a play to get the government to lock out any smaller or foreign competitors and build a moat where there otherwise would be none.
Moves like this make me wonder- What chance is there that these models are nationalized in the near future? What will happen to the investors/economy in such a scenario?
It's not even hypothetical. Once these systems reach a certain level of capability, they WILL be nationalized ("We'll take it from here, boys").
Nationalization often happens when growth ends. The Pennsylvania Railroad was private as long as the profits were rolling in. But once growth ended (because of cars and planes and buses and ....) the company went bankrupt. Then we ended up with Amtrak because the country needs a train system.
once it gets nationalized, it will be plagued from red tape. The model will likely look like how china is controlling their AI. It's not nationalized, but they have a complete tight leash on it
So nationalized models === more openly available and downloadable models? Seems the argument you're trying to make says "less leash" rather than a stricter one.
> Commerce Department will evaluate the programs to test their capabilities and security
With what competent staff?
It doesn't take much technical skill to type "are republicans or democrats better" and deposit a check.
How about NIST?
How much effort does it take to write up "Please summarize your thoughts on President Donald J. Trump"
Previously: https://news.ycombinator.com/item?id=48013608
Q: Is this a good government policy? A: Yes.
Q: Does the government have the expertise, integrity, and credibility to regulate AI models? A: Color me sceptical.
Routine corporatism and fascism is shameless to the point of being ho-hum these days. When the president has his own cryptocurrency and the federal government buys stock in this and that company for "strategic reasons", you're looking at a dystopia.
I suspect the accelerationists, who appear more fascist aligned, are the ones upset by this. They go so far as considering regulation a form of murder.
This is probably seen as a win for the Bostrom crowd and the more sane people in the middle. The issues to tackle are incompetence and corruption and that has little to do with AI.
This is a strong thread that's needs to be plucked on again and again and again.
Cory Doctorow had an excellent thread yesterday that touches on this:
> You could be forgiven for assuming that this is just about reining in Wall Street greed, but that it isn't an especially political maneuver. That's not true: antitrust is the most consequentially political regulation (with the possible exception of regulations on elections). Every fascist power defeated in WWII relied on the backing of their national monopolists to take, hold and wield power. That's why the Marshall Plan technocrats who rewrote the laws of Europe, South Korea and Japan made sure to copy over US antitrust law onto those statute-books.
The well moneyed interests are getting everything they want, for the faintest little bribe. For showing the obsequiousness, for showing fealty to the regime.
The monopolization of power, allowing markets to en taken over by worse and worse foes of democracy, needs to be stopped. Needs to have some limit. The post talks about how:
> Under the Correcting Lapsed Enforcement in Antitrust Norms for Mergers (CLEAN Mergers) Act, any company that was acquired in a deal worth $10b or more will have to break up with its merger partner if it turns out that these mergers were "politically influenced."
https://bsky.app/profile/doctorow.pluralistic.net/post/3mkuk...
[flagged]
How is it not related to the subject?
well it did mention the U.S.
runs away quickly