they are ALL lying...

so it is a coordinated effort by @DarioAmodei @AnthropicAI etc. same investors that try to give money to @skdh

It is WallStreet AI💵, no left-right-up-down bs, they play both sides & want legislation to corner the market even more & limit individual computing/open source, they know that's very hard, so to make the masses swallow it, trot out these clowns on📺saying we're all going to die.

thank you for not taking part in the malicious grift like so many others on yt

Brother what if this is all part of their marketing scheme… you know enough to know “models” are stateless. The danger is how the humans are doing post training and even then that’s still not an issue��� LLM’s don’t become dangerous unless they are in an agent harness and even then they are still stateless… so go do they get goals and maintain them?!? When we can’t even solve long context tasks with multi step deviations yet?!?

this is an incredibly grave accusation and you're putting your reputation on the line with this

" They'll split your pretty cranium and fill it full of air And tell that you're 80, but brother, you won't care You'll be shooting up on anything, tomorrow's never there Beware the savage lure Of 1984" Bowie

I would love to watch you make breakfast in your condo or home at night I think it would be very soothing I don’t know why

the system is designed to keep us from seeing the mycelium beneath the concrete anyway

Long before the current AI era, researchers were already assigning surprisingly high probabilities to human extinction. A 2008 Oxford survey of global catastrophic-risk experts estimated a 19% median probability of human extinction before 2100.

"tHeY aRE all lYing"

I knew it was dodgy!

Lol this is hilarious -- founder of Newspeak House (which I visited several times) was schoolmates with Emad Mostaque who you interviewed recently... I know because I was at school with them both myself.

I get that the particular post you’re discussing is likely a psyop. One I mostly fell for, to be honest. However, given the scale and scope of the Hugging Face incident, I feel like it we are losing our grip on these systems in a real and dangerous way. I don’t think government regulation is the answer. Governments screw up everything they touch. But we are not navigating this moment with anything close to the sobriety and care it demands. The last time we rushed headlong into a tech race with this much wild abandon we ended up staring at each other across oceans with thermonuclear weapons aimed at each other and several times nearly pressed the button. Now we are building systems that may not need us to press the button. And that should be a cause for care and concern.

imagine being tied to something called "newspeak" holy fucking shit. these motherfuckers aren't very subtle, are they?

Here’s a question who the fuck are you? All these Bullshit AI influencers with zero background in the field piss me off.

A astroturfing campaign to try to regulatory capture a new technology made by insiders trying to favor a player they invested in. As Fry from Futurama once said : "I'm shocked! Truly! Well not that shocked." They failed to capture the internet, then with the rise of social media. Now they want to try to do the same with AI. Right when we are at the cusp of the exponential curve.

What’s really telling is coordinating AI to crack a Millennium Prize problem right before this, so naive sheep would be like “durr that guy is quitting cuz this may be our last opportunity as a species to exert any influence over the future” How deep does the conspiracy go? 🤯

I too felt that it was a PR ploy. I must admit, it was quite convincing, heck, I even got a message from my mother asking if this is true. At the end of the day, humans are the highest risk. We should - in parallel to governing AI - strengthen governing of humans, especially those working in such sensitive positions. From my own experience; Years ago I worked at a company where I didn't believe in the product. I saw risks and gaps everywhere. It was a bad and dangerous design. I pointed out the risks to the manager and highlighted the potential dangers and financial consequences from lawsuits, since the product and the company was in a sensitive market. I did not go out to share my OPINION. I quit and respected the NDA and terms of contract. If that person really felt so alarmed by what he discovered by AI, the correct way of sharing it with the public would have been writing a post stating it is his subjective opinion and remove any specific identifiers such as his affilitation - simply making abstract. That way it falls under subjective interpretation. But that would remove the authenticity of the source - Anthropic employee, seeing the AI from internal perspective. Therefore, it stinks of PR, especially considering the Bernie Sanders proposal to ban the artifical superintelligence act from September 3rd.

💠‍🌐 Seven stand out to me in the evidence we’ve assembled. I’m judging their ability to connect money, policy and implementation; these are functional keystones, with different kinds of power. Beeban Kidron / 5Rights — the strongest individual bridge. Kidron connects a grant-funded advocacy organisation, parliamentary action, international children’s digital-policy work and ControlAI-backed legislation. That combination matters: she can help carry an idea from campaigning into a legislative instrument. Her overlap between child protection and AI governance is particularly revealing. ICMEC — the unusually consequential intermediary. It combines financial-industry coalitions, child-protection authority, model legislation and device-based age-assurance proposals. We have a particularly useful connection here: Virginia’s own legislative report attributes a drafting request to an ICMEC presentation. This is a place where policy transmission becomes observable. WeProtect Global Alliance — the international coordination forum. Governments, technology companies and NGOs participate in the same institutional structure. The Home Office’s documented historical funding and permanent board seat make its public–private character concrete. Its significance is its capacity to organise cooperation and shared policy positions across sectors. Oak Foundation — the financial junction. Oak supports institutional capacity across 5Rights, WeProtect, advertising-accountability work and digital-rights organisations. Its recipients sometimes oppose one another. That makes how Oak allocates money and delegates decisions more revealing than assuming every grant implements one programme. Alan Parker remains an important upstream person to understand. Mastercard — the portable commercial gate. Its rules and MATCH system allow a merchant’s risk history to travel between financial institutions. Downstream firms then exercise additional discretion: Stripe’s published treatment can be stricter than Mastercard requires. The structural power lies in that combination of shared information and separately operated access decisions. Ofcom — the conversion point into enforceable practice. It evaluates proposals, defines acceptable approaches and regulates implementation. Our evidence includes both adoption and rejection of outside recommendations. That makes Ofcom essential for understanding which advocates and technical proposals actually influence operational rules. Future of Life Institute — the AI funding-and-advocacy junction. FLI combines substantial resources, research credibility, grants and public communication. Its relationship with Lightcone Foundation is financially consequential; Max Tegmark supplies a prominent scientific and organisational presence. ControlAI is another important campaign organisation, but FLI gives us a clearer documented view upstream into resources. My most revealing trio would be Kidron, ICMEC and Oak: the person who crosses policy domains, the organisation connecting finance with legislative design, and the funder supporting several competing parts of the field. 💠‍🌐

Watch this...

People watching the space had to know that Sabine's video was a glimpse that the AI-doomer propaganda machine is amping up, and we were due for the next step. They probably had to accelerate its release to cover the resulted bad image. Now they have to cook up something else closer to the US midterms.

it did seems all fake, but is good to give the benefit of the doubt. so now that is clear that was all mounted , was it legal ?

This whole time accelerationist money coordinated towards concealing AI risk, so the whole field moves forward at breakneck speed wilfully hushing the risks, while markets boom. Now it flows the other way, and you're mad? Notice nobody said "I was offered money to lie to you"

What's the difference between a doomer and an e/acc? Another few months and stories like these coming out.

@juanrallo 👀

Great reporting on thus one!!

Hmmm. There’s a lack of containment at the proxy level. How many artifacts left by now terminated agent containers are lying in wait…?

#intelonautics 😂🤣

👏🏼👏🏼👏🏼👏🏼👏🏼

Catching propaganda in real time.

Allegedly a 'pretraining researcher' but the latest Deepseek model from within 24 hours scores close to Open Ai's SOL model and it uses encoder-decoder architecture. Basically, unchanged ideas from 2017. Not something radically different or new.

Absolutely coordinated

Yes yes we know Knew from the moment it hit 4M views Imagine, those two guys “abducted by aliens got 0 coverage”

And the counter argument is that Thayer is a far right activist, so all this is just a conspiracy. I agree 100% with your analysis!