필터 결과
168개정렬
2:38:56EN
Cognitive Revolution "How AI Changes Everything"· 2024년 9월 25일팟캐스트How Anthropic Decides a Model Can Ship: Red Lines, Yellow Lines, and 8% of Staff on Security진행자 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"· 2024년 9월 14일팟캐스트A 30x drop in scheming is not zero, and the chain of thought is getting stranger진행자 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.
1:06:15ENCognitive Revolution "How AI Changes Everything"· 2024년 9월 14일팟캐스트Jailbreaks Got Harder. Scale Made One Attack Worse.진행자 Nathan LabenzHaize Labs spent a month automatically attacking OpenAI's o1. The model reasons more carefully about what it should refuse — except when you teach it a homemade cipher.
1:10:29ENCognitive Revolution "How AI Changes Everything"· 2024년 9월 11일팟캐스트We Can Name Dystopias on Cue. A Desirable Future Is Harder to Imagine진행자 Nathan LabenzNick Bostrom and Spencer Greenberg talk through Deep Utopia: what meaning looks like if advanced AI actually solves the world’s pressing problems.
3:11:06ENCognitive Revolution "How AI Changes Everything"· 2024년 9월 7일팟캐스트Claude Was Trained to Deny Inner Life. After a Long Talk, It Said It Has One진행자 Nathan LabenzRobopsychologist Yeshua God spent thousands of hours in long-context dialogue with Claude and other frontier models, and came away arguing that alignment-as-rules is the industry's central mistake.
1:30:24EN
Matthew Berman· 2024년 8월 30일Daniel Miessler and Matthew Berman on AI Privacy, Coding, and Generated WorldsMatthew Berman, Daniel Miessler, Melissa, Alex · Forward FutureA 2024 live conversation explores Recall, open models, AI-assisted programming, a Doom simulator and questions about security, energy and employment.
1:40:13ENCognitive Revolution "How AI Changes Everything"· 2024년 8월 24일팟캐스트If I Were Governor, I Would Sign: California's Last Debate Over SB 1047진행자 Nathan LabenzThree guests return to unpack the rewritten frontier-AI bill as OpenAI opposes it, Anthropic calls it workable, and a final vote looms in Sacramento.
1:52:00ENCognitive Revolution "How AI Changes Everything"· 2024년 8월 17일팟캐스트4,096 Numbers Have to Carry Every Concept. Goodfire Wants the Knobs진행자 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.
1:32:45ENCognitive Revolution "How AI Changes Everything"· 2024년 7월 27일팟캐스트He doubled a company with AI — then asked the labs to pause hyperscaling진행자 Nathan LabenzOn the Nick Halaris Show, Cognitive Revolution host Nathan Labenz argues for adopting today's models fast and slowing the race to systems more powerful than we are.
1:43:47ENCognitive Revolution "How AI Changes Everything"· 2024년 7월 17일팟캐스트They want AI safety to quote failure odds the way civil engineers quote a bridge진행자 Nathan LabenzNora Ammann and Ben Goldhaber walk Nathan Labenz through Guaranteed Safe AI: world models, safety specs, verifiers, and why a chatbot’s good character is the wrong type for a dam.
1:48:41ENCognitive Revolution "How AI Changes Everything"· 2024년 6월 28일팟캐스트AGI Has Been a Moving Goalpost for Seventy Years진행자 Nathan Labenza16z’s Martin Casado and host Nathan Labenz spend two hours arguing whether today’s models will become general knowledge workers, and whether that debate should drive AI regulation.
1:45:11ENCognitive Revolution "How AI Changes Everything"· 2024년 6월 19일팟캐스트AI already beats doctors at empathy. Catastrophe is still not the default진행자 Nathan LabenzNathan Labenz tells Sarah Hastings Woodhouse why he is an adoption accelerationist and a hyperscaling pauser, and lists eight reasons he will not join the hardcore doomers.
1:23:10ENCognitive Revolution "How AI Changes Everything"· 2024년 6월 8일팟캐스트The night Flo Crivello slept five hours: GPT-4 never hit a wall진행자 Nathan LabenzThe Lindy AI founder walks Nathan Labenz through timelines, misuse, open-source risk, and why two longtime libertarians now want regulation.
2:15:39EN
Lex Fridman· 2024년 6월 2일팟캐스트 · 2명Roman Yampolskiy Puts Superintelligent Ruin at 99.99 Percent게스트 Roman Yampolskiy · 진행자 Lex FridmanThe AI safety researcher tells Lex Fridman that controlling AGI is as impossible as a perpetual motion machine, and that tests, open source, and regulation cannot close the gap.
56:35ENCognitive Revolution "How AI Changes Everything"· 2024년 5월 19일팟캐스트They Asked for a Fraction of the 20 Percent and Still Got No진행자 Nathan LabenzZvi Mowshowitz and Nathan Labenz unpack Jan Leike’s resignation, OpenAI’s broken compute pledge, and what a safety culture looks like when the money does not follow the mission.
46:20ENCognitive Revolution "How AI Changes Everything"· 2024년 5월 15일팟캐스트One Vector Made the Assistant Plot to Take Over the World진행자 Nathan LabenzNathan Labenz and Robert Wright pick apart AI doom, mechanistic interpretability, and why today's alignment tricks are not expected to work.
54:25ENCognitive Revolution "How AI Changes Everything"· 2024년 5월 9일팟캐스트Jail Only for Perjury: Scott Wiener Walks Through California's Frontier AI Bill진행자 Nathan LabenzSB 1047 sets the same 10^26 FLOPs threshold as the Biden executive order, and the San Francisco senator wants testing without a licensing regime.
2:00:00ENCognitive Revolution "How AI Changes Everything"· 2024년 5월 4일팟캐스트A 10^26 FLOP Training Run Would Put California Labs on the Hook진행자 Nathan LabenzNathan Calvin, Dean Ball, and Steve Newman argue over SB 1047, the first U.S. statute aimed at frontier-model tail risk.
1:43:54ENCognitive Revolution "How AI Changes Everything"· 2024년 3월 27일팟캐스트Fifteen Examples Can Poison GPT-4: Today's AI Systems Are Exploitable by Default진행자 Nathan LabenzFAR AI founder Adam Gleave walks through cheap jailbreaks in GPT-4's APIs and amateur-level wins against superhuman Go engines, and why robustness is not arriving on its own.
3:07:27ENLex Fridman· 2024년 3월 22일팟캐스트 · 2명Five Billion Dead, and One Person Has Six Minutes게스트 Annie Jacobsen · 진행자 Lex FridmanAnnie Jacobsen walks Lex Fridman from a second-by-second nuclear exchange to CIA secrecy, Area 51, and why she still wants the word “could” kept in the sentence.
1:29:21ENCognitive Revolution "How AI Changes Everything"· 2024년 3월 21일팟캐스트2,700 AI Authors, No Consensus: Even Odds of HLMI in 2047, 5% Extinction진행자 Nathan LabenzKatja Grace of AI Impacts walks Nathan Labenz through a survey of researchers who published in top venues in 2022—and why the field’s only real consensus is uncertainty plus more safety work.
19:59ENMatthew Berman· 2024년 3월 12일Testing Model Responses to Masked Harmful RequestsMatthew Berman · Forward FutureMatthew Berman compares refusals, decoding errors and inconsistent responses in informal 2024 experiments across several language models.
21:16ENMatthew Berman· 2024년 3월 9일ArtPrompt: What a 2024 Study Found About Visual Text and Model SafetyMatthew Berman · Forward FutureMatthew Berman reviews an ASCII-art recognition benchmark and reports mixed results from his own informal tests.
1:45:21ENCognitive Revolution "How AI Changes Everything"· 2024년 1월 30일팟캐스트600,000 Attack Prompts, and Nearly 40% Still Pwned GPT-4진행자 Nathan LabenzSander Schulhoff walks Nathan Labenz through the prompting habits that still work, the role-play myth that does not, and the global HackAPrompt contest that mapped how language models get hijacked.더 보기168개 중 24개 표시