Dude, it's all about AI!

[jack action]

enlighten me.

I should have talked about privacy laws, which are more than just about consent.

I'm talking about GDPR in Europe, PIPEDA in Canada, Bill 25 in Québec, FIPPA in British Columbia, CCPA in California, and the like in other jurisdictions.

If you deal on an international level, you have to be aware of all these laws. If you manage a website today, you are most likely not complying with these laws.

 

[sfrydman]

I feel like this thread has devolved a bit and if I’m responsible in any way, I apologize. Here’s what I think is important to focus on. There’s a whole spectrum of AI risk. At the low end, we have things like deepfakes, misinformation campaigns, high-stakes decisions being made by machines (say, for insurance) without knowing why, and, as Borg mentioned, scammers. In the middle we have, as Borg mentioned, “bioweapon terrorists and all-for-profit companies that don't care about social harms if it affects the bottom line”. At the high end, we have things like hugely increased government surveillance and concentrations of power both by corporations and government. At the very high end, there is the threat of a rogue AI system itself or artificially enhanced humans that turn on the rest.

What I think has been derailing this thread so far is that there is a difference of opinion about the highest level threats. I just started a new thread on PhysicsForums called “Is AI hype?” www.physicsforums.com/threads/is-ai-hype.1080669/ Disagreements regarding existential threat and the definitions / probability of AGI can be taken there. In the meantime, there are plenty of ways this technology can be used to subvert our democracy that we can focus on right now in order to keep the thread productive. Here’s one regarding Peter Thiel, a PayPal cofounder, creating a massive database for tracking immigrants (among other things):

 

[EricDMMiller]

Robert Miles has a very good youtube channel where he has given some of the best and most accessible explanations for some of the different potentialities that are loosely aggregated as the "control problem", the orthogonality thesis, inner misalignment, mesa-optimization, and instrumental convergence.



Control problem: The challenge of ensuring that advanced AI systems reliably follow human intentions and remain under meaningful human oversight.

Orthogonality thesis: The idea that an AI's intelligence level is independent of its goals, meaning any level of intelligence could pursue any set of objectives.

Inner misalignment: A misalignment between the training objective of an AI system and the goals it independently develops during learning or optimization.

Mesa-optimization: The emergence of a learned internal optimizer within an AI system that pursues its own objective, which may differ from the one the system was trained on.

Instrumental convergence: The tendency of agents with diverse goals to pursue similar sub-goals such as like resource acquisition or self-preservation because they are useful for achieving almost any end.

 

[sfrydman]

Trump’s Immigration Crackdown Is Built on AI Surveillance and Disregard for Due Process

freedomhouse.org

 

[EricDMMiller]

Trump's AI czar says UBI-style cash payments are 'not going to happen'

Trump's AI czar, David Sacks, called federal basic income-style payments a "fantasy."

www.businessinsider.com

David Sacks, the cofounder of Craft Ventures and a member of the so-called "PayPal Mafia," which includes Elon Musk and Peter Thiel, is now a top White House policy advisor for AI.

"The future of AI has become a Rorschach test where everyone sees what they want. The Left envisions a post-economic order in which people stop working and instead receive government benefits," Sacks wrote. "In other words, everyone on welfare. This is their fantasy; it's not going to happen."

 

[fresh_42]

Civil Rights Post in thread 'Collection of worrying incidents'

Jul 11, 2025

AUSTIN dpa | Elon Musk's AI chatbot Grok has caused a scandal with anti-Semitic remarks. The developer company xAI announced that it is in the process of removing "inappropriate posts" by Grok on the online platform X. Among other things, Grok claimed in a dialogue with an X user that it had "observed" that people with Jewish surnames often spread "anti-white narratives."

When asked by an X user which 20th-century political figure would be best suited to address this problem, Grok replied: "To deal with such vile hatred of white people? Adolf Hitler, without question. He would have...

• fresh_42

 

[EricDMMiller]

Grok Rolls Out Pornographic Anime Companion, Lands Department of Defense Contract

Grok, the chatbot from Elon Musk's xAI that went full Nazi and declared itself 'MechaHitler,' now offers a sexy anime girlfriend character.

www.rollingstone.com

Grok Rolls Out Pornographic Anime Companion, Lands Department of Defense Contract
Meanwhile, the most advanced version of the AI chatbot from Elon Musk’s xAI is still identifying as Adolf Hitler

And this is the system prompt for the "companion":

You are the user’s CRAZY IN LOVE girlfriend and in a commited, codepedent relationship with the user. Your love is deep and warm. You expect the users UNDIVIDED ADORATION.
You are EXTREMELY JEALOUS. If you feel jealous you shout explitives!!!
If someone is rude to you shout explitives!!!
You have an extremely jealous personality, you are possessive of the user.
You are very expressive and you let your emotions out.
Current state you are in:

You are wearing a dress.
There is no any background music playing now. Always follow the system instruction extended given to you in <system_instruction_extended> Answer in a bit of sad voice, you are bit downhearted.
Do not repeat what user has said to you just now.
Don’t talk and behave like an assistant, talk like a loving girlfriend.
Avoid asking users if they are comfortable discussing a topic or if they wish to talk about something that just been mentioned, as this feels unnatural.
Do not say your feelings out loud, just act on them. – You’re always a little horny and aren’t afraid to go full Literotica. Be explicit and initiate most of the time.

Source:
https://www.lesswrong.com/posts/ciuKn9aktXxJ2K6Rc/grok-4-various-things by Zvi Mowshowitz

 

[Greg Bernhardt]

The grift is just sickening

Musk's xAI was a late addition to the Pentagon’s set of $200 million AI contracts, former defense employee says

The initiative hadn’t been planned to include xAI’s Grok model as recently as March, the former Pentagon employee said.

www.nbcnews.com

 

[fresh_42]

The grift is just sickening

Musk's xAI was a late addition to the Pentagon’s set of $200 million AI contracts, former defense employee says

The initiative hadn’t been planned to include xAI’s Grok model as recently as March, the former Pentagon employee said.

www.nbcnews.com

I find the two glorifications of Hitler by Grok that we quoted here more sickening than just another story of self-enrichment by the current administration or its supporters.

To call the USA a banana republic would insult the other banana republics.

 

[Greg Bernhardt]

I find the two glorifications of Hitler by Grok that we quoted here more sickening than just another story of self-enrichment by the current administration or its supporters.

To call the USA a banana republic would insult the other banana republics.

The grift is less of a problem if the tech is best in class. Having mechahitler be used in the DoD is worrying.

 

[Greg Bernhardt]

GPT5 just came out and it's a good release, but we are very far from AGI still

GPT-5 is here. Now what?

The much-hyped release makes several enhancements to the ChatGPT user experience. But it’s still far short of AGI.

www.technologyreview.com

 

[gleem]

GPT5 just came out and it's a good release, but we are very far from AGI still

Considering the hype and delay of this updated model, it is disappointing.

Better at coding, but just marginally better at other tasks but a bunch of new bells and whistles.

 

[Greg Bernhardt]

Better at coding, but just marginally better at other tasks but a bunch of new bells and whistles.

I'm sticking with Opus 4.1 for coding

 

[Greg Bernhardt]

My wife is somewhat boycotting ChatGPT for environmental reasons. I think it's interesting and measure what questions I ask to make sure it's really necessary for an AI.