Trump orders US government to cut ties with Anthropic Just Hours Before Deadline

[Tumblin Tumbleweed]

Where is Congress on all this?

The dysfunctional morons don't realize that autonomous weapons bypass their controls on war?

The Pandora's box is open. You voted for this. Embrace the suck.

 

[Anomalism]

No. I don't like this was handled.

What are your thoughts on the president's comments about Anthropic?

 

[TemplarKormac]

What are your thoughts on the president's comments about Anthropic?

Hmm. Not much. Just that I disagree with how it was handled. It amounted to nothing but a tantrum.

 

[Anomalism]

Hmm. Not much. Just that I disagree with how it was handled. It amounted to nothing but a tantrum.

I feel like we've all seen this movie before.

 

[TemplarKormac]

I feel like we've all seen this movie before.

Yes we have.

 

[MisterBeale]

Well that's the thing about value; it's subjective.

It matters to me. That's why I asked.

As an independent observer of the process, I'm always on the side of the process, not the policy.
This is the only guarantee that America as a nation and culture can continue.




For instance, would you support someone who politically is your opposite, someone who you don't really agree on about anything, if you knew they were honest and had integrity. . .

. . or are you the type that would support someone whose politics you mostly agree with but whom you know is corrupt as the day is long?

 

[RhodyPatriot]

View attachment 1225386

Anthropic had a $200 million Pentagon contract. They were already the first AI company running models on classified military networks. They were in. Then Defense Secretary Hegseth issued a memo requiring all Pentagon AI contracts to include "any lawful use" language, meaning the government could use the AI however it wanted as long as it fit their interpretation of being technically legal.

Anthropic drew two specific lines...

No mass domestic surveillance of American citizens.

No fully autonomous weapons, meaning AI that selects and fires on targets with zero human involvement.

Those were their terms from the original contract. The Pentagon wanted them removed. Anthropic said no. Trump's response was to threaten civil and criminal consequences and direct every federal agency to immediately cease using their technology. He called them radical left woke nutjobs who have no idea what the real world is about.

Then OpenAI signed the contract within hours. Here's what I want you to actually think about...

Mass surveillance of American civilians. Does that sound like something you want an AI system doing without contractual restrictions? The Fourth Amendment exists for a reason. "Trust us, it's already illegal" is what every government says before it does the thing. There's also already precedent for the government spying on us.

Autonomous weapons with no human in the loop. A system that selects and eliminates targets without a human making the final call. On American soil potentially. Against American citizens potentially. The President is angry that a private company won't build that without restrictions.

Trump's statement doesn't engage with any of this. It's pure dominance framing. "That decision belongs to your Commander in Chief." Not "Here's why surveillance restrictions are wrong." Not "Here's why autonomous weapons need no oversight." Just "I decide. Submit or face consequences."

The company that refused is now being punished with the full power of the executive branch. The company that complied got the contract. You can think Anthropic is a left wing company. You can dislike their politics on other issues. But on these two specific things, surveillance of American citizens and autonomous weapons without human control, they drew a line that any consistent constitutionalist should respect.

Do you work for Anthropic?

You've been obsessed about this to the point of starting multiple threads.

 

[scruffy]

What are your values specifically?

I communicate those to my elected representatives. That's the way this works, right?

Are you not in alignment with the idea that AI shouldn't be used for mass surveillance on Americans?

It already is! Jeez... I can't say what I know. But I can point you someplace. Admiral John Q Poindexter. Remember him?

At the time he got busted, he was running something called OTIA, the Office of Total Information Awareness.

And the country considered him so vital, so important, that two years after he got convicted for perjury he had a job again and he was leading the team again.

What about AI that can autonomously decide to kill people?

I seriously don't think the DoD will invent any such thing. The last thing they want is a war getting out of control and turning on themselves.

What we really have to worry about are the rogue elements, the cartels and terrorists of the world. They have access to more or less the same technology the DoD does.

Do you agree with the president's comments about Anthropic?

Not sure I heard them. You mean, he called them woke?

That's just Trump being Trump. I don't pay much attention to what he says, I try to stay focused on what he does.

I agree with the president that the military needs the technology. And I agree that Anthropic isn't the only player in the game. The correct way for Anthropic to handle this would have been to request a security clearance like everyone else. This game has been going on for a long, long time. Experienced DoD contractors like Northrop and Harris know how to handle it.

 

[Anomalism]

Do you work for Anthropic?

You've been obsessed about this to the point of starting multiple threads.

It's important.

 

[scruffy]

It's important.

Well, let's be practical.

Congress can make a law, and the only people that'll break it are the bad guys. (Think, gun laws - same deal).

Congress can then ask the President to enforce, which policy is already in play for things like hacking (FBI), trans-national hacking (CIA), and state sponsored military hacking (DoD). I worked in cyber security for a number of years. Keeping up with the hackers is an endless task. If they start using AI and we can't, we're screwed.

You might be surprised at the things that become part of the national infrastructure. Earthquake sensors, and rainfall sensors. Offshore sonobouy's. All kinds of oddball stuff.

So, vigilance is a key concept. It's not the same as surveillance, but it's close in some cases. If we get hacked, we can put the AI to work figuring out how they did it, but a better thing to do would be to have an AI intrusion detection system, that automatically backtracks the source. It would have to surveil the network, which is to say filter everyone's input. But that's already being done, if you have a network provider you're going through at least three layers of it.

I mean, everyone and their brother is already surveiling you! They're selling your information to whoever'll buy it, you can buy it online, they don't ask who you are. If the government wanted your info they could buy it too. (Well, they get it for free by subpoena).

We're already immersed in surveillance, there's not a whole lot we can do about it. The other thing, about autonomous weapons, is the reason we have surveillance. One guy, with a bad attitude or a suicide wish, could destroy the world with an autonomous weapon. That's reality, today. No corporation and no government is going to stop him, unless they make the parts impossible to get (which is impossible, 'cause there's always someone out for a profit), or unless they can catch him in the act of building it. Which... y'know... purchase history, I mean the FBI is looking at all this stuff already, and they're using AI to do it. So is Google, IRS, the banks... everyone and their brother.

 

[scruffy]

Interesting quote from this article:

• Anthropic's artificial intelligence models are being used to support the U.S. military's operations in Iran, even after the company was blacklisted by the Trump administration, a source told CNBC.

Iran war live updates: Trump says U.S. to offer insurance for Gulf shipping and escort tankers

Follow the latest news after the U.S. and Israel launched a massive attack on Iran that killed Iran's Supreme Leader Ayatollah Ali Khamenei.

www.cnbc.com

What does this tell you? "Models"? They're writing code on the fly?

The only thing that makes sense is they're retraining AI's based on data from the field. They're using Anthropic's "workflow".

 

[ThunderKiss1965]

The reason I'm posting this despite all of the trashing of my original attempt to make others aware of what is happening, is because this is important.

There are people here on this message board who think that an AI model is nothing more than a glorified search engine, yet the military is so pissed off at the company who created Claude, Antropic, that's when it couldn't coerce them into letting them exploit their intellectual property and use it for potentially nefarious purposes, Trump ordered the severance of all contracts with them.

MSN

That's how that works with Government contracts moron. If a company can not or will not fulfill a request the Government can terminate the contract.

I have serious concerns about Government surveillance and overreach no matter who is in charge but you loons seem to turn that off and own based on the party running the show.

 

[Anomalism]

That's how that works with Government contracts moron. If a company can not or will not fulfill a request the Government can terminate the contract.

I have serious concerns about Government surveillance and overreach no matter who is in charge but you loons seem to turn that off and own based on the party running the show.

Right.

Them dropping Anthropic is the obvious move for their interests.

Let's focus on their interests though. The government wants the ability to create mass surveillance AI and AI that can autonomously decide to kill people.

Thoughts?

 

[ThunderKiss1965]

Right.

Them dropping Anthropic is the obvious move for their interests.

Let's focus on their interests though. The government wants the ability to create mass surveillance AI and AI that can autonomously decide to kill people.

Thoughts?

I made my thoughts clear on Government surveillance.

 

[scruffy]

Apparently ChatGPT users are having a conniption.

Sam Altman in Damage Control Mode as ChatGPT Users Are Mass Cancelling Subscriptions Because OpenAI Is "Training a War Machine"

Sam Altman was blasted in an embarrassing Q&A while scores of ChatGPT users ditch the service in favor of Anthropic.

futurism.com

 

[NewsVine_Mariyam]

Principle #1: when you buy something, it's yours. You can do whatever you want with it. Upgrade it, throw it away, connect it to something else... whatever you want

You couldn't be more wrong. When you "buy" a software license, it generally states you're purchasing the right to use or utilize the software. You don't OWN it.

You don't have to believe me, try reading some of the what you agreed to when before you were allowed to install it.

 

[Anomalism]

Apparently ChatGPT users are having a conniption.

Sam Altman in Damage Control Mode as ChatGPT Users Are Mass Cancelling Subscriptions Because OpenAI Is "Training a War Machine"

Sam Altman was blasted in an embarrassing Q&A while scores of ChatGPT users ditch the service in favor of Anthropic.

futurism.com

Claude's paid membership has doubled in like a month.

 

[NewsVine_Mariyam]

Interesting quote from this article:

• Anthropic's artificial intelligence models are being used to support the U.S. military's operations in Iran, even after the company was blacklisted by the Trump administration, a source told CNBC.

Iran war live updates: Trump says U.S. to offer insurance for Gulf shipping and escort tankers

Follow the latest news after the U.S. and Israel launched a massive attack on Iran that killed Iran's Supreme Leader Ayatollah Ali Khamenei.

www.cnbc.com

What does this tell you? "Models"? They're writing code on the fly?

The only thing that makes sense is they're retraining AI's based on data from the field. They're using Anthropic's "workflow".

I would suspect they're simply violating the terms of agreement & licensing.