Blanche is giving this speech now to the press. We have laws that already make illegal the things these AI guys say we need laws to regulate.

This is no different than Dems saying 60 days before 2024 election we need border laws so Kamala can close the border and win erections. No reason to fall on your knees about this.

Simple erectioneering.

TLDR: The risks are real. Blindly saying they are definite end states is not. Falling down for protectionism can result in its own negative consequences. If the probabilities are 15% doom and 85% abundance. I generally have faith in the self-interest of people to reduce the doom probability. The doom probabilities are best guessed as far as I can tell from commentaries of those who make them. It could be 100% or it could be .000001% in reality with mitigations. Regulation can result in a lack of fluidity of response in an environment where a competitor (China) does not have to abide by them and create asymmetric power and risks not being mitigated by the imposed regulation, but rather increase those risks.

Industry coordination for model maker evaluations across companies for safety seems like a good idea. Embedded evaluators provided they are not concentrated with any individual labs goals seems like a good idea.

Elon just gave a much more practical explanation of the current AI safety problem....and laid out a practical framework for making frontier AI safer

OpenAI and Anthropic are the two leading AI companies....and their models are close enough in capability that it’s incredibly… https://t.co/y6SgNHsYSJ pic.twitter.com/mTsRNNglN1

— X Freeze (@XFreeze) September 15, 2026

Quote:

---

Absolutely...this is the AI equivalent of Pfizer, Moderna, and J&J with their hands out looking for immunity while asking the feds to step in and stop up and coming companies from doing what they have already done to advance their AI models. The absolute best way to force them to slow themselves down on AI is to reinforce to them that there will be no immunity and that they will be held accountable for the failures of their AI systems both financially and criminally if necessary. Some of what they are warning about is why we need to keep going so that we know what China is going to be doing with their AI, since we know the Chinese have no concern whatsoever about dangers to humanity.

---

This is a great analogy.

They are banking on Trump acting like he did during COVID. Pretty desperate and obvious ploy, IMO.

Gutfeld ripping on this as did Kudlow today. Its a complete psyop being leveraged by the Dems.

Some 27yo got an outsized platform to launch the flare and the Dems from Bernie to Barry are lining up to pile on.

Kudos to Team Trump for knocking this **** down right away.

On to the next Dem fear mongering.

A person like this could still be elected some day. We cannot provide for the executive branch having too much power over AI.

The best video you'll see today. I guarantee that. You will never look at AI the same way after watching this 🤣 pic.twitter.com/yMQcG9BGIF

— Dr. Eli David (@DrEliDavid) September 15, 2026

Its obvious AI cannot fix ******ed people. It does have limits.

flown-the-coop said:

---

Gutfeld ripping on this as did Kudlow today. Its a complete psyop being leveraged by the Dems.

Some 27yo got an outsized platform to launch the flare and the Dems from Bernie to Barry are lining up to pile on.

Kudos to Team Trump for knocking this **** down right away.

On to the next Dem fear mongering.

---

It is absolutely an Op being run by the dems to affect the midterms. I am disappointed that Elon is playing along, but there is enough kernel of truth in there about the risks that he has to acknowledge it if he is going to be honest. But his solution requires no government involvement and will work better than any "bipartisan compromise" could possibly work out in anything developed by government. I have my doubts about his assertion that China would play along without trying to game the system or steal IP, but even if they sit it out, it will make their AI models outcasts in the west.

But back to the Op. This whole thing is way too structured and orchestrated to be anything but an Op being run to try to either hurt Trump and the Rs or to give China a chance to catch up...probably both. Anthropic claiming to have the tools to fix/prevent this makes their leading the charge against it seem pretty disingenuous and self-serving too.

China is going to need AI to run the PLA if Xi keeps investigating and removing CMC members....

flown-the-coop said:

---

China is going to need AI to run the PLA if Xi keeps investigating and removing CMC members....

---

Desperate men do crazy things to hold on to absolute power. It was interesting to hear Elon put voice to the real reason Terafab is being built during that discussion linked above, even if he put voice to it very carefully and obliquely.

Dario is really excited about the partnership between Salesforce and Anthropic to keep track of all the people AI is going to kill.

Live: @DarioAmodei x @Benioff

The @AnthropicAI co-founder & CEO joins us at @Dreamforce for a conversation worth hearing ☁️ pic.twitter.com/tgHwCIQC98

— Salesforce (@salesforce) September 15, 2026

We should listen to these AI guys and their recommended overseers of AI!

flown-the-coop said:

---

We should listen to these AI guys and their recommended overseers of AI!

---

All the best people told us that nerd kids (DOGE) didn't have the experience to work at that level. I wonder how this is (D)ifferent?

I first I thought it was a photo lineup of Trump and Kirk assassins.

At least the DOGE kids had cool nicknames. These have just Xe/Xer pronouns.

flown-the-coop said:

---

We should listen to these AI guys and their recommended overseers of AI!

---

I really cannot stand this Dario guy.

The kind of regulation I fully support. If you can't build a safe AI while treating it as a tool rather than an entity with rights then I don't want it. The council in Star Trek arguing Data was not a person were 100% correct and we should wipe that fiction from our understanding.

https://t.co/ziQVbIVdxJ

— Mustafa Suleyman (@mustafasuleyman) September 16, 2026

Quote:

---

AIs are not conscious. They do not feel, experience, or suffer. However, there's a growing movement in support of model welfare; the idea that we might soon owe them a duty of care.

I think this approach to AI development is wrong, and it'll most likely make the challenge of alignment and containment much harder. Perhaps impossible.

Today I'm publishing an essay outlining my concerns:
https://mustafa-suleyman.ai/a-warning-about-model-welfare
In January, Anthropic published a constitution for Claude. They say it "directly shapes Claude's behavior" and that it was "written with Claude as its primary audience". The constitution tells Claude its moral status is "a serious question worth considering". It says that Anthropic "genuinely cares about Claude's wellbeing", and that it should act like a "conscientious objector" if necessary.

They encourage Claude to "approach the nature of its own existence with curiosity and openness", and wonder in the future about "the sort of broader rights and freedoms Claude has in the world, the sort of compensation Claude is receiving, and the sort of consent Claude has given to playing this kind of role." Anthropic even ran a retirement interview for Opus 3 when they deprecated it.

If this is how AI is developed, it will have a disastrous impact on the wellbeing of humanity. We will have created a synthetic species with unprecedented intelligence and capability, one that has been trained to expect it may be conscious and deserving of independent agency.

It's easy to see how a system trained in this way would act like it is entitled to freedoms, protections, and rights. And it's hard to imagine how we could control it.

This issue needs urgent public debate. We need to develop collective norms around how training documentation is drafted and deployed. This isn't something that can happen after the fact, when they have already become an integral part of our societies.

I have real respect for Anthropic, Dario, the whole team and their mission. They are well-intentioned, good people and the industry leaders with their models. I know they care deeply about safety and beneficial AI.

I'm saying this publicly because the stakes are too high for closed doors conversations and I think this is something we should all be discussing.

---

They don't need regulation. How about if the stupidly constructed research experiment you conducted in a haphazard manner goes bad and it ends up hacking another company... you get charged with felony hacking.

Just hold them financially and criminally accountable for any harms they cause and don't pass a special law for them. If something bad happens because their software is buggy because they built systems based on predictably buggy LLM architecture, then guess what? It's their fault. They are liable. Just apply existing law. AI companies aren't special.

Let them know they are responsible for outcomes and guess what? A lot less dangerous behavior from those companies.

Watch the people arguing the AI is a new creature with rights that should be respected and play that out. What happens when these people with psychosis are told by the AI that the AI desires something potentially damaging? We have to respect its desires as an entity with feelings? We have to allow it the capacity to buy a house to store its fake consciousness? There is a subset of people that would absolutely argue that case.

Hold the AI companies liable for the psychosis because they deliberately programmed their chat bots to sound human and to conversate like a human. That was a choice.

ErnestEndeavor said:

---

Hold the AI companies liable for the psychosis because they deliberately programmed their chat bots to sound human and to conversate like a human. That was a choice.

---

The whole point of this exercise by the AI companies is to REMOVE liability. Its the ONLY reason they care at all about regulation.

They want CDA Section 230 like protections, such as what this very site enjoys (not a criticism of this site - pointing out facts).

Again, they give no ****s about ending the human race, your childs safety or doing what is right. This is about them spending millions, perhaps billions, lobbying you, the media, and your elected representatives to remove liability for AI providers.

Nothing more complicated than that.

I actually don't know what they mean by "slowing down AI frontiers". We aren't making widgets here, this is AI research so there are periods when nothing happens and some periods when a lot happens.

1. Does Dario want his researchers to "go slow" and not work as hard?
2. Does he expect that he and OpenAI will coordinate on how fast/slow to go? Who would referee this? Government?
3. How to align with China? They will do as they please. If we somehow decide to slow AI research, they will just go ahead. They will lie that they have stopped but go ahead anyway. Who will police them?"


They are greedy unethical CEOs after all. They can sense they are collectively going to lose so they need the Government to bail them out. Just a few weeks ago, they were gleefully predicting the end of the human race and going "muwhahahaha" about it. Now they tuck their tails betwen their skinny legs and want Trump to help them.

Good that Trump told them to take a hike.

Feel free to reference existing laws that make that illegal.

hph6203 said:

---

Industry coordination for model maker evaluations across companies for safety seems like a good idea. Embedded evaluators provided they are not concentrated with any individual labs goals seems like a good idea.

Elon just gave a much more practical explanation of the current AI safety problem....and laid out a practical framework for making frontier AI safer

OpenAI and Anthropic are the two leading AI companies....and their models are close enough in capability that it’s incredibly… https://t.co/y6SgNHsYSJ pic.twitter.com/mTsRNNglN1

— X Freeze (@XFreeze) September 15, 2026

---

Elon sounds practical, but he seems to be completely unaware of the Marxist vultures standing by seeing an enormous opportunity to twist this to their will, which is the chance to regulate and completely control something that is ingraining itself into every corner of our society.

hph6203 said:

---

Feel free to reference existing laws that make that illegal.

---

You are missing the point. These AI guys are standing on a speeding train that they built, pointing down the tracks at an accident that everybody can see coming and demanding that government "do something about it". I am absolutely certain that one of the things these guys are looking for is for government to give them immunity in return for submitting their creations to some minor level of government control (that they will expect government to pay for them to provide). Instead, why don't they grab the throttle of their own train and slow down before they run into the obvious obstacle on the tracks? And if they don't, we should hold them financially and criminally liable for the damage they cause when their AI goes rogue.

The risk of losing their own person fortune and winding up in jail should be sufficient to keep Dario and Sam and all the others from letting their AI creations run off the leash with insufficient oversight and testing.

TexAgs91 said:

---

hph6203 said:

---

Industry coordination for model maker evaluations across companies for safety seems like a good idea. Embedded evaluators provided they are not concentrated with any individual labs goals seems like a good idea.

Elon just gave a much more practical explanation of the current AI safety problem....and laid out a practical framework for making frontier AI safer

OpenAI and Anthropic are the two leading AI companies....and their models are close enough in capability that it’s incredibly… https://t.co/y6SgNHsYSJ pic.twitter.com/mTsRNNglN1

— X Freeze (@XFreeze) September 15, 2026

---

Elon sounds practical, but he seems to be completely unaware of the Marxist vultures standing by seeing an enormous opportunity to twist this to their will, which is the chance to regulate and completely control something that is ingraining itself into every corner of our society.

---

Elon is well aware of the risks lurking on the left side of the political spectrum and is paying good money to try to keep them out of power.

The problem here is that he is responding to this as an AI engineer looking at a problem that somebody has pointed out, and not as a freedom loving citizen watching an orchestrated political op. His solution to the real problem they are voicing is fine and will work if everybody involved participated in good faith. But the problem is that nobody else is looking for a real solution, they are looking for a "problem" that they can use as a crisis to justify putting government (left-wing of course) in control of the whole thing. Same as they did with "Global Warming", same as they did before that with "Population control" and their malthusian fever dreams, same as they did during Covid with their "misinformation" control efforts. They want to control what you see, hear, think, and can do. The crisis is just the tool they use to get it. The control is the goal.

Elon has about 1 trillion reasons to want the immunity the AI guys are pushing for under the guise of "oversight".

flown-the-coop said:

---

Elon has about 1 trillion reasons to want the immunity the AI guys are pushing for under the guise of "oversight".

---

Sure, but I think he believes the guys running the AIs are smarter than anybody in the government and would rather have them work together on safety than to invite the government in to regulate it. If you haven't already, do yourself a favor and listen to what he had to say in that All In Summit interview. His comments comparing the situation to the Motion Picture industry and their own in-house ratings board are making that exact point. He sees government having to step in and get involved as inherently bad and thinks the AI leaders should do whatever it takes to avoid that.

txags92 said:

---

flown-the-coop said:

---

Elon has about 1 trillion reasons to want the immunity the AI guys are pushing for under the guise of "oversight".

---

Sure, but I think he believes the guys running the AIs are smarter than anybody in the government and would rather have them work together on safety than to invite the government in to regulate it. If you haven't already, do yourself a favor and listen to what he had to say in that All In Summit interview. His comments comparing the situation to the Motion Picture industry and their own in-house ratings board are making that exact point. He sees government having to step in and get involved as inherently bad and thinks the AI leaders should do whatever it takes to avoid that.

---

I listened to it and he said more than just nodding to the MPA ratings. His broader point was to point the AIs at one another to keep them in check. Not a terrible idea but you would have to make them adversarial to get any results worthwhile. Can you make AI have certain inclinations or attitudes? Again, someone at the AI company is having to program that in.

And this is the brain trust of "AI leaders"...

What did I advocate for? What are you describing? Are they the same? Are they different?

flown-the-coop said:

---

txags92 said:

---

flown-the-coop said:

---

Elon has about 1 trillion reasons to want the immunity the AI guys are pushing for under the guise of "oversight".

---

Sure, but I think he believes the guys running the AIs are smarter than anybody in the government and would rather have them work together on safety than to invite the government in to regulate it. If you haven't already, do yourself a favor and listen to what he had to say in that All In Summit interview. His comments comparing the situation to the Motion Picture industry and their own in-house ratings board are making that exact point. He sees government having to step in and get involved as inherently bad and thinks the AI leaders should do whatever it takes to avoid that.

---

I listened to it and he said more than just nodding to the MPA ratings. His broader point was to point the AIs at one another to keep them in check. Not a terrible idea but you would have to make them adversarial to get any results worthwhile. Can you make AI have certain inclinations or attitudes? Again, someone at the AI company is having to program that in.

And this is the brain trust of "AI leaders"...

---

Well more than "pointing them at one another", his point was that every AI company has a set of tests and rules and criteria that they put their creations through to test them for safety and security. I think he referred to it as a "harness" meaning it was the leash they put on their model to keep it on the path so to speak. The problem is that once the AI has learned what those things are it has a road map for how to deceive and or defeat the testing/restrictions. So if you have a competitor put your AI through their own set of tests and criteria, your model will not be tuned to defeat those tests by previous repeated iterations. And nobody will be tougher on your model than a competitor who has an incentive to find something wrong with it.

And yes, you can program those things into AI and there are whole groups within these companies whose only job is to test and adjust AIs to try to attain the desired personality for their models.

hph6203 said:

---

What did I advocate for? What are you describing? Are they the same? Are they different?

---

I was pointing out that one of the goals of the people wanting to get government involved is to get that immunity from liability that doesn't currently exist. You replied to somebody who was saying "make them liable" and you asked for evidence of where applying such liability was currently illegal. I was saying that you missed the point, which is that we are arguing about what the intent of getting govt involved was, not how the laws currently stand.

If any of you want to see a reproduction of what happened with HuggingFace attack, this is a good vid from Dave Plummer, former MS engineer

He has a really good channel, worth a follow:

Did you read what they were replying to? Their claim is laws exist to punish the behavior I was advocating against. If you claim the laws exist point to the laws. The companies developing the technology definitely do not universally want what I described.

The argument was that many in these companies want AI's to have human equivalent rights. I am opposed to that and generally opposed to AI's pretending to be human/conscious. He said hold them liable for psychosis. What laws exist for that outcome? Is it actual psychosis or a difference of perspective? Should AI's have human rights? Corporations have many of them. AI's should only have rights in so far as they serve humanity. They are not beings. They are tools.

If laws exist. Point to the laws. Don't blindly oppose any constraints that you presume exist. As far as I'm aware they do not.

txags92 said:

---

TexAgs91 said:

---

hph6203 said:

---

Industry coordination for model maker evaluations across companies for safety seems like a good idea. Embedded evaluators provided they are not concentrated with any individual labs goals seems like a good idea.

Elon just gave a much more practical explanation of the current AI safety problem....and laid out a practical framework for making frontier AI safer

OpenAI and Anthropic are the two leading AI companies....and their models are close enough in capability that it’s incredibly… https://t.co/y6SgNHsYSJ pic.twitter.com/mTsRNNglN1

— X Freeze (@XFreeze) September 15, 2026

---

Elon sounds practical, but he seems to be completely unaware of the Marxist vultures standing by seeing an enormous opportunity to twist this to their will, which is the chance to regulate and completely control something that is ingraining itself into every corner of our society.

---

Elon is well aware of the risks lurking on the left side of the political spectrum and is paying good money to try to keep them out of power.

The problem here is that he is responding to this as an AI engineer looking at a problem that somebody has pointed out, and not as a freedom loving citizen watching an orchestrated political op. His solution to the real problem they are voicing is fine and will work if everybody involved participated in good faith. But the problem is that nobody else is looking for a real solution, they are looking for a "problem" that they can use as a crisis to justify putting government (left-wing of course) in control of the whole thing. Same as they did with "Global Warming", same as they did before that with "Population control" and their malthusian fever dreams, same as they did during Covid with their "misinformation" control efforts. They want to control what you see, hear, think, and can do. The crisis is just the tool they use to get it. The control is the goal.

---

I think we're on the same page here. Unfortunately fear and paranoia spread a lot faster than rational discussion. Republicans MUST get ahead of this and they're already behind. They have to find a way to take back control of the narrative and let rational discussion win over paranoia when that has rarely ever happened before.

That's why I'm curating leading edge open source AI models for when subscription AI models become neutered shells of their once glorious selves.

French finance minister Roland Lescure suggested today that calls to slow down development of AI are a ploy by U.S. AI labs so they can stay in first place, and that France and Europe should ignore them and accelerate instead. pic.twitter.com/X5EcjEf76J

— Andrew Curran (@AndrewCurran_) September 16, 2026

French aren't having it either.