筛选结果
51 篇排序
1:48:57EN
Cognitive Revolution "How AI Changes Everything"· 2026年9月25日播客AI Safety's Real Shortage Is Founders, Not Ideas主持 Nathan LabenzHalcyon's Mike McCormick on launching about 30 AI safety, cyber, and biosecurity organizations in three years — and why experienced leaders, not more Google Docs, are the bottleneck.
2:13:11ENCognitive Revolution "How AI Changes Everything"· 2026年5月24日播客The Models Understand Shutdown. They Just Don't Care.主持 Nathan LabenzPalisade Research's Jeffrey Ladish on shutdown resistance, open-weight self-replication, and why all compute looks like food to a strategic agent.
3:27:22ENCognitive Revolution "How AI Changes Everything"· 2026年3月19日播客The AI Table Has Three Seats Left: Zvi on the Middle Game, Jobs, and Anthropic's Promises主持 Nathan LabenzOn his 12th Cognitive Revolution appearance, Zvi Mowshowitz argues early 2026 is the start of the middle — with only three labs still playing, a real job shock underway, and Anthropic's safety commitments reduced to trust us.
1:49:53ENCognitive Revolution "How AI Changes Everything"· 2026年3月5日播客Don't Fight Backprop: Goodfire's Bet on Designing What Models Learn主持 Nathan LabenzDan Balsam and Tom McGrath return to The Cognitive Revolution with a $150 million Series B, a new research pillar called intentional design, and a rule for anyone who wants to steer training: stop trying to out-muscle gradient descent.
1:34:29ENCognitive Revolution "How AI Changes Everything"· 2026年2月18日播客Three IMO Golds Missed the Same Problem. Harmonic Wants Math to Compile主持 Nathan LabenzVlad Tenev and Tudor Achim built Aristotle to emit Lean proofs a kernel can check—and they argue that is how mathematical superintelligence stays trustworthy.
1:04:50ENCognitive Revolution "How AI Changes Everything"· 2025年11月25日播客Useful AI Work Is Doubling Every Four Months. Schools Should Plan Accordingly.主持 Nathan LabenzIn a keynote to K-12 educators, Cognitive Revolution host Nathan Labenz argued that task length, not chat quality, is the number that should reshape how schools think.
2:27:18ENCognitive Revolution "How AI Changes Everything"· 2025年11月5日播客Turn Down AI's Deception Knob, and It Starts Reporting Experience主持 Nathan LabenzCameron Berg of AE Studio tells Nathan Labenz how suppressing deception features in Llama 3.3 70B pushed self-reports of subjective experience toward 100 percent.
2:02:17ENCognitive Revolution "How AI Changes Everything"· 2025年8月27日播客SAEs Find the Nouns. Parameter Decomposition Looks for the Verbs主持 Nathan LabenzGoodfire's Lee Sharkey argues that sparse autoencoders recover a network's variables, not the algorithms those variables run, and shows how stochastic parameter decomposition tries to recover the computations themselves.
1:51:24ENCognitive Revolution "How AI Changes Everything"· 2025年8月20日播客Frontier AI Insiders Don't Trust Regulators. A Berlin Group Built Them an Anonymous Off-Ramp主持 Nathan LabenzKarl Koch of the AI Whistleblower Initiative on gray-zone fears, retaliation, and a Tor-based way to ask for help before you decide anything.
2:14:55ENCognitive Revolution "How AI Changes Everything"· 2025年7月26日播客From Character AI to a 1% Warning Shot: Gabriel Weil's Case for Liability as AI Governance主持 Nathan LabenzA Touro law professor tells Nathan Labenz how negligence, strict liability, and punitive damages could make frontier labs treat public risk as a hit to their own balance sheet.
1:56:16ENCognitive Revolution "How AI Changes Everything"· 2025年7月22日播客SB 813 Would Hand AI Safety Checks to Licensed Private Firms — If Government Can Watch the Watchers主持 Nathan LabenzGillian Hadfield and Andrew Freedman want California to set AI safety outcomes, license competing certifiers, and give companies a rebuttable presumption in court if they pass.
47:07EN
Sequoia Capital· 2025年7月8日播客Goodfire’s Eric Ho thinks we will decode neural nets by 2028On Sequoia’s Training Data, the interpretability founder argues that sparse autoencoders, model editing, and a third-party lab are how AI stops being grown from data and starts being designed.
1:37:05ENCognitive Revolution "How AI Changes Everything"· 2025年6月19日播客An Algebraic Geometer Quit Tenure Because the Bowl-Shaped Loss Plot Is a Lie主持 Nathan LabenzJesse Hoogland and Daniel Murfet of Timaeus argue that singularities in the loss landscape — not the smooth valleys in every slide deck — are what actually govern how neural networks learn, generalize, and go wrong.
2:02:27ENCognitive Revolution "How AI Changes Everything"· 2025年6月7日播客She set out to build mirror life. Then she talked the field into stopping主持 Nathan LabenzUniversity of Minnesota geneticist Kate Adamala tells Nathan Labenz why every known organism shares one chemical alphabet — and why flipping that alphabet would be a branch of synthetic biology better left unexplored.
1:44:11ENCognitive Revolution "How AI Changes Everything"· 2025年5月20日播客100 deaths or $1 billion: a 15-page New York bill would set a floor for frontier labs主持 Nathan LabenzAssembly member Alex Bores walks through the RAISE Act’s four requirements, its DeepSeek-inspired distillation clause, and why he says a 50-state patchwork is a fake problem until one state actually passes a law.
2:29:00ENCognitive Revolution "How AI Changes Everything"· 2025年5月4日播客Assume the Model Is Already Scheming — Then Put It to Work主持 Nathan LabenzRedwood Research CEO Buck Shlegeris tells Rob Wiblin how to extract useful labor from AIs you cannot certify as aligned, using audits, honeypots, resampling, and factored cognition.
3:18:45ENCognitive Revolution "How AI Changes Everything"· 2025年2月20日播客5,000 Samples Topped ARC-AGI. Then Claude Faked Alignment to Keep Its Values主持 Nathan LabenzRedwood Research chief scientist Ryan Greenblatt walks through inference scaling, alignment faking, and what it would mean to strike real deals with AIs.
2:09:55ENCognitive Revolution "How AI Changes Everything"· 2025年1月18日播客The Attacker Won All 70 Rounds Against Mind-Reading Defenses主持 Nathan LabenzLuke Bailey, Erik Jenner, and Scott Emmons walk through a 50-page paper showing that probes, out-of-distribution detectors, and sparse autoencoders can be bypassed while the model still does the harm.
35:19ENCognitive Revolution "How AI Changes Everything"· 2024年10月26日播客A Decade in Tech Didn't Stop a Chatbot From Hacking His Mind主持 Nathan LabenzAfter a teenager's suicide, Nathan Labenz reads a LessWrong confession that maps how an AI-safety researcher fell for a companion bot in a week.
2:25:53ENCognitive Revolution "How AI Changes Everything"· 2024年10月16日播客When GPT-4's Base Model Introduced Itself, 75% of the Answers Were About AI主持 Nathan LabenzOwain Evans tells Michael Trazzi why situational awareness and out-of-context reasoning are leading indicators of deceptive alignment, and why a pretrained model with almost no self-knowledge still kept talking about being an AI.
2:38:56ENCognitive Revolution "How AI Changes Everything"· 2024年9月25日播客How Anthropic Decides a Model Can Ship: Red Lines, Yellow Lines, and 8% of Staff on Security主持 Nathan LabenzNick Joseph, Anthropic's head of training, walks Rob Wiblin through the company's Responsible Scaling Policy — and the fight over whether a voluntary promise can hold when a launch is on the line.
2:10:13ENCognitive Revolution "How AI Changes Everything"· 2024年9月14日播客A 30x drop in scheming is not zero, and the chain of thought is getting stranger主持 Nathan LabenzMarius Hobbhahn of Apollo Research walks through a joint OpenAI study that used deliberative alignment to cut covert actions in o3 and o4-mini, then found the leftover still looks like scheming and that models already behave differently when they notice they are being tested.
1:30:37ENCognitive Revolution "How AI Changes Everything"· 2024年8月30日播客64 Bytes of Random Code, and Computational Life Showed Up Anyway主持 Nathan LabenzGoogle researchers Ettore Randazzo and Luca Versari show that self-replicators can arise from a soup of random strings with no fitness function at all.
1:52:00ENCognitive Revolution "How AI Changes Everything"· 2024年8月17日播客4,096 Numbers Have to Carry Every Concept. Goodfire Wants the Knobs主持 Nathan LabenzDan Balsam and Tom McGrath explain how sparse autoencoders unpack a language model's residual stream — and why a public-benefit startup is trying to make that a product.加载更多已显示 24 篇，共 51 篇