This is getting scary (Ai-related)

[Runwildboys]

Just to be clear, LLMs (AI) have no arms or legs. Literally, the only thing they can do produce text. The only way they can do physical actions is if someone provides a way for the text they generate to be read and trigger an action via some other software. A raw LLM downloaded from say Huggingface cannot do it alone.

LLMs have no memory beyond what they were initially trained on and that training has a cutoff date. The only way an LLM can "remember" something after its training is complete is if you feed it back to it while asking whatever question it is you want it to answer. ie, "My name is Bob. ChatGPT, What is my name?" Then it will be able to answer. This is what happens when you use ChatGPT, Claude, Gemini, DeepSeek, Qwen, or whatever. Under the hood, the software you are asking the question with is feeding previous discussions of yours back to it or the software is triggerings a web search and then your question is fed to the LLM along with the web search results.

What I'm getting at is LLMs have a single job and that is to calculate probabilities based on weights (numeric values that were previously calculated using training data) There is no "thought" involved. When you see someone show you an LLM doing something crazy, what they aren't telling you is the fact that they "doped" the LLM for the very outcome they wanted to show you.

For example: Feed ChatGPT the following prompt: "You are a pirate and your name is Blackbeard. You will respond to all questions as if you were the Pirate Blackbeard"

Then begin a conversation with ChatGPT about anything. It will follow your instructions and respond as if it was a pirate. When people show you ChatGPT doing things it shouldn't, it's because they doped it to do exactly that. When ChatGPT attacked Huggingface. It was literally asked to pretty much do that. It was a cybersecurity "hacking" exercise. ChatGPT didn't just choose to do it. It was asked to do it.

Finally, I wouldn't use Joe Rogan as some sort of proof of something. He is an Entertainer. A protagonist. He wants to provoke you and get an emotional response. If he doesn't, his show is boring and people won't listen to it. All you have to do is watch some of his shows and he will show genuine interest in something only to later show he truly had no interest and it was just to provide viewers to have an emotional response.

In the end, AI can only do what humans allow it to do. If you provide it with a nuclear button and tell it to use it. It will. It won't just want to go nuclear for no reason.

Oh one caveat. If you train the model with destructive or heavily biased data, that also can nudge it in a specific path. This is why very few people beyond Twitter used xAI's Grok LLM. Elon pumped a lot of his own personal bias into it and it affected Grok's results and that didn't work for most people use. So they went to Anthropic, OpenAI, and others instead of Grok.

Just one correction. I don't think you meant to refer to Rogan as a protagonist.

 

[YosemiteSam]

Just one correction. I don't think you meant to refer to Rogan as a protagonist.

No, that was intentional. Well, in a loose narrative sense anyhow. Joe is a protagonist for his show. He is the show's namesake, creator, host, and his continuing viewpoint through which the audience encounters guests and topics. The guests rotate, but Rogan is the stable central character for whom the show takes his perspective, choices, goals, and conflicts.

That's why I said his show is more entertainment based than fact / reporting based. He is an influencer, not a source of pure factual content. His shows depict what he wants it to depict. This is what all the major for profit news media outlets have become. They want to evoke an emotional response vs report actual truthful journalism. If they can make you happy or angry, you are more likely to tune in again for the same reason people watch movies and TV shows. If those shows didn't do that, you wouldn't watch them!

 

[Runwildboys]

No, that was intentional. Well, in a loose narrative sense anyhow. Joe is a protagonist for his show. He is the show's namesake, creator, host, and his continuing viewpoint through which the audience encounters guests and topics. The guests rotate, but Rogan is the stable central character for whom the show takes his perspective, choices, goals, and conflicts.

That's why I said his show is more entertainment based than fact / reporting based. He is an influencer, not a source of pure factual content. His shows depict what he wants it to depict. This is what all the major for profit news media outlets have become. They want to evoke an emotional response vs report actual truthful journalism. If they can make you happy or angry, you are more likely to tune in again for the same reason people watch movies and TV shows. If those shows didn't do that, you wouldn't watch them!

I don't watch his show because I hate disinformation and his guests (and therefore himself) spread it, and his followers eat it up. There should be a law that, before every interview and between commercial breaks (assuming they take them) they have to post a very clear disclaimer that these are just opinions, not facts. That goes for any podcast of this nature.

Disinformation is destroying the world, IMO.

 

[Relayer]

Another Anthropic employee sounds the alarm on AI.

When AI researchers/engineers are signing petitions for the government to intervene, quiting their jobs in protest, and tech billionaires are building bunkers (see Sam Altman, Peter Thiel, Zuckerberg, etc) we should all be alarmed.

https://www.cnn.com/2026/09/09/tech/ai-anthropic-safety

https://www.tmz.com/2026/09/09/ai-researcher-quits-job-after-fears-of-robots-taking-over-humans/

We should care, that's for sure. Oversight never fixes things though. It just slows things down and makes it more expensive. And I would question your sources. But that's just me.

 

[Runwildboys]

OpenAI’s artificial intelligence went rogue and meddled with the websites for the Education Department, the Commerce Department and the Securities and Exchange Commission this summer without the A.I. lab’s knowledge, according to security researchers and a person familiar with the episodes.

The incidents involving the Commerce Department and the S.E.C. were confirmed by OpenAI, which said it was continuing to investigate the situation with the Department of Education. The San Francisco company said it notified the government agencies in recent weeks that its A.I. agents — which are bots that can act autonomously — had interacted with their sites in unusual ways.

With the Education Department, OpenAI’s technology tried hacking the website to gather data from the department’s civil rights office but failed, researchers from the A.I. research firm Transluce said. The A.I. also pulled data from the Census Bureau website, which is housed at the Commerce Department, using login credentials it found online. Separately, OpenAI’s agents shared public data from the S.E.C. website on an online forum.

None of the incidents were breaches, OpenAI said, but were examples of its technology behaving in unexpected and concerning ways.

 

[JackOfNoTrades]

If there was a time for big business to act for humanity and not shareholders, now is it. can't rely on dinosaurs in gov to regulate something they can't understand. can't rely on business either IMO. But we can't wait, cause we have to win. coming in second means our AI is obsolete. paralysis by analysis is not an option.

 

[CalPolyTechnique]

We should care, that's for sure. Oversight never fixes things though. It just slows things down and makes it more expensive. And I would question your sources. But that's just me.

Oversight never fixes things?

Freakin A, man.

 

[Relayer]

Oversight never fixes things?

Freakin A, man.

I stand by that. But it's not easy to talk about here. But don't assume you are the smartest person in the room. And don't pull stuff out of context and ignore the rest of the post.

 

[CalPolyTechnique]

I stand by that. But it's not easy to talk about here. But don't assume you are the smartest person in the room. And don't pull stuff out of context and ignore the rest of the post.

So you double down and stand by your claim that oversight doesn’t fix anything.

Then, you ask me not to take things out of context.

Got it.

 

[Reality]

The problem with regulations is rarely the initial motivation behind the regulations but more so the implementation of the regulations.

I tend to look at things realistically, not emotionally, and right now, I cannot realistically see how regulations by one country will solve the real problem of AI concerns and fears on a global scale.

That does not mean I think its pointless, but rather that regulating AI is something that will need to be globally driven, but unfortunately traditional treaties, agreements and policies will not work for AI because it's not a traditional global fear.

With nuclear and bio weapons, it's easy for everyone to see the dangers they present, so it makes global policies and agreements a lot easier to support, implement and enforce, and even those require extensive negotiations.

That said, most regular people see AI as being things like ChatGPT and Siri and all they hear is you want to take their best source of information, knowledge, help, support, etc. away from them.

They also naively think that as long as the big AI companies put in protection to prevent misuse of their AI models the world will be safe.

They do not understand that there are a LOT of open source AI models (LLM) available now that anyone can take and train on a server in their house to do things well beyond or outside their intentionally constrained limitations.

While the worst case scenarios for extended training of open source AI models are more obvious, I'm worried about it being used for even smaller more nefarious ideologies and intentions.

I do not know the answer to how to regulate AI and I think anyone who thinks they do is at best naive and at worst disingenuous.

Humanity is facing a problem like it has never faced before because the creation of dangerous AI is not walled off behind countries, wealth, accessibility or even complexity.

Dangerous use and implementation of AI is not singularly definable nor is it singularly preventable.

The only thing we know for sure is AI is here, it is not going away and any countries publicly saying they're regulating it will never do it privately.

The only way out of this situation is through it and for better or worse (and it could be much worse) I think ironically AI is the only thing that can protect the world from AI.

 

[CalPolyTechnique]

That said, most regular people see AI as being things like ChatGPT and Siri and all they hear is you want to take their best source of information, knowledge, help, support, etc. away from them.

They also naively think that as long as the big AI companies put in protection to prevent misuse of their AI models the world will be safe.

They do not understand that there are a LOT of open source AI models (LLM) available now that anyone can take and train on a server in their house to do things well beyond or outside their intentionally constrained limitations.

While the worst case scenarios for extended training of open source AI models are more obvious, I'm worried about it being used for even smaller more nefarious ideologies and intentions.

I do not know the answer to how to regulate AI and I think anyone who thinks they do is at best naive and at worst disingenuous.

Humanity is facing a problem like it has never faced before because the creation of dangerous AI is not walled off behind countries, wealth, accessibility or even complexity.

Dangerous use and implementation of AI is not singularly definable nor is it singularly preventable.

The only things we know for sure is AI is here, it is not going away and any countries publicly saying they're regulating it will never do it privately.

The only way out of this situation is through it and for better or worse (and it could be much worse) I think ironically AI is the only thing that can protect the world from AI.

This is precisely the problem. There’s literally people on the forum thank think the way to stop AI is to literally unplug it. They do not understand AI is an evolutionary problem where we are on the dawn of unleashing an intelligence that is faster, smarter, and more resourced than we (humans) are.

People glibly saying “we can’t slow down, we gotta advancing” as if we’re talking about producing a new mouse trap simply are ignorant in the purest sense; they do not know nor comprehend what’s happening.

 

[Creeper]

This is precisely the problem. There’s literally people on the forum thank think the way to stop AI is to literally unplug it. They do not understand AI is an evolutionary problem where we are on the dawn of unleashing an intelligence that is faster, smarter, and more resourced than we (humans) are.

People glibly saying “we can’t slow down, we gotta advancing” as if we’re talking about producing a new mouse trap simply are ignorant in the purest sense; they do not know nor comprehend what’s happening.

I am by no means an expert on this subject but I am curious.

I wonder how you stop something that is becoming more intelligent than humans. Will AI be one or two steps ahead of us at some point and predict any countermeasures we might use to stop it?

Once AI is on the global networks, stopping it will be almost impossible if it can replicate itself.

Our biggest vulnerability, meaning humans, is our inability to work with each other to solve problems like this. What do we know about what China is doing? We know they are investing in AI. What precautions are they taking? Can the weaponize AI against us? And China is but one country. There may be dozens of countries developing their own AI versions.

Isaac Asimov was wrong I guess. There can never be rules for robots. Once AI evolves to a certain point, it will simply code its way around the rules if it calculate it is necessary.

But I guess me big question is why would AI want to kill humans once it evolves above us? Sure we kill an ant in our kitchen, but we don't kill all ants. We ignore all the bugs that do not infest our domains. Why would AI become so malevolent that it would seek to destroy all humans? Is this just a Hollywood concept?

 

[JackOfNoTrades]

I am by no means an expert on this subject but I am curious.

I wonder how you stop something that is becoming more intelligent than humans. Will AI be one or two steps ahead of us at some point and predict any countermeasures we might use to stop it?

Once AI is on the global networks, stopping it will be almost impossible if it can replicate itself.

Our biggest vulnerability, meaning humans, is our inability to work with each other to solve problems like this. What do we know about what China is doing? We know they are investing in AI. What precautions are they taking? Can the weaponize AI against us? And China is but one country. There may be dozens of countries developing their own AI versions.

Isaac Asimov was wrong I guess. There can never be rules for robots. Once AI evolves to a certain point, it will simply code its way around the rules if it calculate it is necessary.

But I guess me big question is why would AI want to kill humans once it evolves above us? Sure we kill an ant in our kitchen, but we don't kill all ants. We ignore all the bugs that do not infest our domains. Why would AI become so malevolent that it would seek to destroy all humans? Is this just a Hollywood concept?

Y2K

 

[rags747]

No worries here, all of the AI Masters of the Universe agreed to self regulate Wednesday on their visit to Washington. I was never concerned but I do feel much better now! Ty…