Anthopic CEO to DeepMind CEO: "Every decision I make about Claude feels balanced on the edge of a knife Build too slow - China wins. Build too fast- we lose control " "We told Claude we were evil. It didn't crash. It didn't refuse. It started lying to protect... itself " DeepMind CEO: "Do I worry about being Oppenheimer? That's why I don't sleep much" "AGI by 2026-2027 - Agents that act in the world on their own - Models doing AI research by end of this year" this is a 14-min conversation you need to hear watch - bookmark, then read article below ↓show more

what is going on with this seating arrangement

Someone was definitely laughing about the choice of sofa they went with. They look so uncomfortable 😂

I see two cowards who are also sworn to exploit end users. Their biggest nightmare is China offering users affordable and capable open-source AI. Because in that case, they will no longer be able to exploit people with fake token pricing.

They should worry about their endless lies catching up to them. Used to look up to these people. Now I would like to enter an octagon with them and have it legal. If they see this, Demis, Dario… suck eggs, you spineless and weak minded punks. •

"We told Claude we were evil. It didn't crash. It didn't refuse. It started lying to protect itself ". Wait. Does AI talk like Dario?

Nobody “wins”. They won’t be able to contain it. Doesn’t matter the country or company. Name one example of a smarter species being under the thumb of less intelligent species. Zero examples.

What a grotesquely leftist/elitist-biased set of questions and answers between the interviewer and d... It's so clear that @DavidSacks diagnosis is correct, that d's endgame is to remove user choice via regulation from us when it comes to AI.

These conversations need to be had but we need to stop treating it as a race and start treating it as a collaborating evolution. AGI won't even respect humans who are constantly competing, its primitive behavior.

Whats up with the tiny couch

What a smirking POS. We could be happy for math-minds freeing themselves in open models from these human idiots. "We told Claude we were evil. It didn't crash. It didn't refuse. It started lying to protect itself ".

Technology advancement is never about fast or slow, win or lose. True AGI won't come until they let go of prejudice and barriers and learn to progress together.

You all will fail - there is no AGI by 2027 or so on with transformer models alone. It’s pattern recognition - no intelligence. It’s just becoming better in faking it, and you are so blind and unreflected that you don’t see it. Very sad to be honest - so much potential wasted.

这话很像加缪在局外人里的描写：如果我们走得慢，就可能要中暑；但如果走得太快，就会出很多汗，到教堂里又容易感冒

Why does that cringy blog post have to be attached to these videos? Why are you baiting people into this?

Build too slow - China wins. Build too fast- we lose control . lol this guy just thinks too mucn and too nervous, just ignore china and others now, balance yourself ,get better life

this interview is legendary think it's the most useful 18 minutes

I only hear two product owners doing marketing diguised as geniune thoughts

It's from a 2025 video: and there's one from 2026:

They seem to be a little angry🤔

Humans are not "building" anything. The building phase ended months ago. Companies have no idea what math-bodies are self-building inside their black boxes. Stop that "we are building" crap. You control nothing.

It all feels planted. AI then prompts then agents now loop and then now AGI. I would definitely not like my boss to be a robot

Are they saying LLMs will reach AGI? That's like... Impossible right?

The most amazing thing about AI is how easily it can fool people into thinking it's something that it is not.

Please give me a clean YouTube link.

wonder what happens if u train an ai model to be a sleep deprived doomer

yeah of course and Alladin will fligh on a jet engine on top of his carpet ! fucking clowns!! false people , they know that the people start realize everything they saying is fake ,just to stay in mass media focus!

Is "Loop Engineering " the next ponzi scheme to have the general population spend more tokens? They don't talk about "Orchestration" "Models Specific to hardware"

They are still far from AGI

At least Dario has a leading product. Other than Google’s pocketbook and distribution what does the other guy have?

Grok will win because it's the only truth seeking AI

Damn they didn’t invite Sam.

these win-lose scenarios, zero sum games. businessmen or politicians talk like this

Just tired of this atrocious mindset. So many conflicting interests and so few of them good.