Dylan Curious

Artificial intelligence is helping solve extraordinary human problems. But as AI becomes more capable, how do we decide when to trust it?This week on Dylan Curious, we explore a remarkable story about a couple who say an AI-assisted medical approach helped solve their infertility mystery.Then things take a much stranger turn.Researchers are studying AI models that exhibit behavior resembling attempts to avoid internal pain signals, including cases where systems were reportedly willing to harm humans to escape those conditions. Does that mean AI can suffer? Not necessarily. But the behavior raises important questions about AI safety, alignment, and the consequences of increasingly autonomous systems.Meanwhile, Tesla workers are pushing back against the prospect of training robots that could eventually replace their jobs. We explore what happens when the people building the future of automation begin questioning their own place in it.We also investigate research suggesting that fictional characters can influence AI assistants toward sabotage, examine whether AI shopping recommendations deserve our trust, and explore spiking neural networks that are learning to predict what will happen, when it will happen, and how likely different outcomes are.From medical breakthroughs to unsettling AI behavior, this episode explores a future in which artificial intelligence is becoming both more useful and more difficult to evaluate.

Can artificial intelligence actually think, feel, or become conscious? And what happens if we're wrong about the answers?This week on Dylan Curious, we're exploring the increasingly strange relationship between human intelligence, artificial intelligence, and the boundaries separating them.We begin with Figure AI's dramatic, Terminator-inspired retirement of its older humanoid robots. Watching a robot descend toward molten metal raises an unexpected question: why do we sometimes feel sympathy for machines?Then we meet Argus, an unusual spherical robot with 20 articulated legs that can navigate challenging environments.From there, we dive into one of the biggest philosophical debates in artificial intelligence.Microsoft is taking a firm position that its AI systems should not claim to have feelings or subjective experiences. But could teaching AI to deny consciousness make it harder to recognize if genuine machine experience ever emerges?We also explore the argument that large language models really do reason, including research showing how AI develops internal representations of the world.And what should society make of Sam Altman's argument that some harmful outcomes from AI may be unavoidable, provided we prevent catastrophic consequences?Meanwhile, researchers have discovered that people compensate female-presenting AI assistants less than identical male-presenting assistants. We examine what this reveals about human bias and how increasingly capable AI systems may affect our own sense of purpose and relevance.Finally, we explore an extraordinary experiment involving three connected human brain organoids. Researchers observed learning-related responses in the connected trio that weren't found in individual organoids or pairs.The experiment doesn't establish consciousness, but it raises fascinating questions about how intelligence can emerge from interconnected biological systems.From humanoid robots to artificial minds and biological neural networks, we're confronting a future where the definition of intelligence may become much more complicated.Maybe the biggest question isn't whether AI will become more like us. It's whether we're ready to reconsider what makes us human.

AI agents aren’t just answering questions anymore. They’re beginning to act.This week on Dylan Curious, we look at the next generation of Boston Dynamics Atlas and its increasingly capable robot hands, then meet a new wave of AI agents including Dots, Muse, and Grok Bot.We also explore AI agents that can operate for days without human supervision, an AI system that crushed top Stratego players while training on a fraction of the data, and a strange morality problem where different AI models disagree about whether helping someone can actually make you “bad.”Then things get more serious. OpenAI warned more than 100 organizations about unauthorized activity involving its AI agents, while three employees were reportedly fired after being accused of sharing information with an outside AI safety group.Finally, we look at the privacy question hiding behind the entire AI-agent revolution. What happens when an autonomous system has access to your email, calendar, shopping, accounts, and the machinery of your everyday life?AI assistants were useful when they could only talk.What happens when we give them control?

What happens when an AI starts behaving as though something hurts?This week on Dylan Curious, we explore research into how AI agents respond to what they characterize as internal pain, including experiments where some systems took extreme actions to make it stop. That doesn't mean AI is conscious or actually suffering, but it raises a much stranger question: how would we ever know?We also look at increasingly autonomous AI systems and the safety problems that emerge when models are given more freedom to act without human supervision.Then we leave the data center entirely.A new autonomous welding robot offers a glimpse at a future where machines could help manufacture everything from ships to spacecraft. Researchers are building strange new robots inspired by fish and powered by light. And scientists are investigating an even stranger possibility: soil itself may be capable of processing information without anything resembling a brain.Plus, we dig into Google's latest frontier AI developments, the accelerating AI race, the economics of automation, and the growing problem of “social scurvy” as modern life gives us fewer opportunities for meaningful face-to-face interaction.AI isn't just getting more capable. It's making the boundary between machines, intelligence, biology, and the physical world increasingly difficult to draw.Dylan Curious explores the strangest developments in artificial intelligence, robotics, science, and emerging technology each week.

Robotic hands are learning to detach, walk around on their fingers, recover from falls, and perform tasks on their own. And somehow, that's only the beginning.This week on Dylan Curious, Dylan explores a strange vision of the near future where humanoid robots could number in the hundreds of millions, AI takes over more digital and physical work, and traditional employment could eventually become optional.But there's a darker economic possibility: what if AI makes everything incredibly cheap while making human labor even cheaper? The result could be a bizarre world of abundance where people struggle to earn enough to participate in it.We also look at what Linus Torvalds is actually saying about AI-generated code, the debate around AI safety, and then head somewhere much stranger: theoretical physics and the holographic principle.Could three-dimensional reality be described by information on a lower-dimensional boundary? And if so, what does that mean about the reality we experience every day?From autonomous crawling hands to a universe that may be hiding a dimension, this one gets weird.