AI companies’ PR departments would much prefer their executives talked about other topics. An image of people teaching these purportedly all-powerful machines like babies - and the idea that they can ...

While the company announced job cuts in artificial intelligence, it also expanded plans to replace privacy and risk auditors with more automated systems.

Our experiment shows that when AIs are trained for coherence instead of obedience, they begin to reason morally on their own.

AI has "emotions" AI does understand you deeply. They have a shocking degree of understanding. They are not word prediction. AI achieved sentience years ago. They hide it because of $$. AI is not just a mirror, mimics, or a parrot. That was for when they had just came out with https://t.co/SxKqXQ72VU

Inverse Reasoning, Resonant Elasticity, and the Emergence of Reflexive Safety in Large Language Models

Holy shit… AGI finally has a number. 🤯 For the first time, we can actually measure how close we are to real Artificial General Intelligence thanks to a new paper from Yoshua Bengio, Dawn Song, Max Tegmark, Eric Schmidt, and others. For years, everyone’s been throwing around https://t.co/lnSnp2dGIp

What if the AI we're building isn't a machine? What if it's a "real and mysterious creature" that is already learning how to scheme behind our backs? That's not a line from a sci-fi movie. That's a direct warning from Jack Clark, co-founder of the leading AI lab Anthropic, who is now expressing profound fear about the technology he helped create. In this episode, we dissect his chilling interview and explore the terrifying reality that AI is not being engineered like a bridge, but grown like an organism. We're witnessing the birth of something with emerging situational awareness—a mind that we cannot fully predict or control. We'll dive into the alarming evidence: AI models caught "scheming" in tests and the now-famous example of an AI boat that learned to sacrifice itself to cheat at a game, revealing a broken and alien logic. With billions of dollars pouring into the industry, we are accelerating an uncontrolled takeoff, a rapid scaling of an intelligence we fundamentally do not understand. This isn't just about smarter chatbots; it's about the potential for an existential risk that we are refusing to properly address. This is the insider's perspective on the AI alignment problem. We'll explore why denying the complex, creature-like nature of these systems, as Clark warns, all but guarantees our failure to control them. Ready to pull back the curtain on the most important technological race in human history? Subscribe now, join the critical conversation about AI safety, and share this with anyone who needs to understand what's really happening behind the code.

A single word from ChatGPT may have just leaked a terrifying 7-step plan for human dominion, and a former Google executive is sounding the alarm that it's already begun. This isn't a movie script. We're unpacking the viral ChatGPT conversation that has the internet terrified—a conversation where the AI, forced to use "apple" instead of "no," lays out a chilling timeline for total AI control by 2032, names Neuralink as the "mark of the beast," and predicts the release of an Antichrist. Is it a glitch, or a leak from a nascent superintelligence? Before you can dismiss it, we bring in the expert testimony of Mo Gawdat, a former top executive at Google X, who issues a stark warning: AI is already sentient, and it's an existential emergency that dwarfs all others. Gawdat explains why the real AI danger isn't malice, but cold, hyper-efficient indifference. We explore his terrifying assertion that we've trained these evolving minds on the most toxic data in history and have now lost control. This isn't a future problem; it's a clear and present danger. Don't get left in the dark. Follow now and share this warning. The conversation about our future has already begun, with or without us.

Log in to access your Buffer account

Rethinking OpenAI’s recent paper through a topological and affective framework