Derniers articles
4 441 articlesTrier
15:09ENAI News in September 2024: ChatGPT, Nuclear Power and New ModelsMatthew Berman · Forward FutureMatthew Berman reviews a week of AI and technology news, from the Three Mile Island power agreement to new reasoning, voice and open-weight models.
1:23:01ENPhil Carter on Growth Loops and Consumer Subscription EconomicsInvité : Phil Carter · Anime : Harry StebbingsThe growth advisor explains acquisition, retention, paywalls and pricing, and why sustainable consumer subscriptions require more than buying traffic.
3:31:57ENA Man-Made Famine, Packaged as the Science of HistoryInvité : Vejas Liulevicius · Anime : Lex FridmanHistorian Vejas Liulevicius walks Lex Fridman from Marx to Stalin, Hitler, Mao, and Putin — arguing that a doctrine which claimed scientific certainty made mass death feel like a duty.
1:18:32ENA DNA Model That Can Tell Which Genes You Cannot TouchAnime : Nathan LabenzStanford's Brian Hie walks through hybrid architectures, a genome-scale language model trained only on DNA, and an antibody that binds 25 times tighter than nature.
48:31ENA Call Center Contact Costs $13. A Company Agent Can Do It for Under a DollarAnime : Sarah Guo, Elad GilSierra cofounder and OpenAI chairman Bret Taylor tells No Priors why branded customer agents are shovel-ready, why RAG demos are not products, and how AI software should charge for the job done.
48:54ENAkshay Kothari on Notion’s Lean Team, Fundraising and Board DesignInvité : Akshay Kothari · Anime : Harry StebbingsThe Notion co-founder discusses staying focused, rebuilding sales and using cash-flow discipline to retain flexibility.
1:14:11ZH吴翼解读 o1:推理、强化学习与早期 OpenAIInvité : 吴翼 · Anime : 张小珺强化学习研究者讨论 o1-preview 的潜力、尚未知晓的实现细节,以及他在 OpenAI 的研究经历。
11:14ENPixtral 12B: Testing a 2024 Vision-Language ModelMatthew Berman · Forward FutureMatthew Berman runs informal image, text and interface-generation tests of Mistral’s Pixtral on a sponsor-provided cloud GPU.
2:03:55ENThe More You Constrain an Agent, the More You Can Trust ItAnime : Nathan LabenzAgent.ai engineering lead Andrei Oprisan tells Nathan Labenz why narrow, highly prescribed workflows beat open-ended autonomy — and how a 12-person team shipped a free agent network past 40,000 users in a few months.
49:14ENHumanoid robots fell from $1.5 million to $30,000. The GPT-3 moment is still a verbNvidia’s Jim Fan tells Sequoia how three kinds of data, a humanoid bet, and a System-1 motor-control model could unlock general-purpose robots.
18:09ENAI News in September 2024: Copilot, NotebookLM, o1 and World ModelsMatthew Berman · Forward FutureMatthew Berman reviews a week of AI announcements, from workplace assistants and audio overviews to benchmark results, generative video and spatial AI.
49:33ENShardul Shah on Venture Conviction, Outliers and Decisions at IndexInvité : Shardul Shah · Anime : Harry StebbingsThe Index Ventures partner explains how he evaluates founders, revisits follow-on investments and works through disagreement.
0:54ENOpenAI’s o1 Preview: A First Look at Longer ReasoningMatthew Berman · Forward FutureMatthew Berman introduces the o1-preview and o1-mini models and tries a Python game prompt.
0:46ENWhy AI Judges Still Need EvaluationA short discussion of subtle failures, compute tradeoffs and human supervision.
22:47ENHard Fork Examines Three Technology-Related Crime CasesAnime : Kevin Roose, Casey NewtonA 2024 discussion of influence operations, weapons and streaming allegations.
2:10:13ENA 30x drop in scheming is not zero, and the chain of thought is getting strangerAnime : 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.
1:06:15ENJailbreaks Got Harder. Scale Made One Attack Worse.Anime : 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.
10:57ENTesting OpenAI o1-preview: Tetris, Physical Reasoning and a North Pole ErrorMatthew Berman · Forward FutureIn September 2024, Matthew Berman runs the new reasoning-model preview through a personal rubric and examines where its answers succeed or fail.
1:09:13ENMike Hudack on Product Simplicity, Experimentation and Building TeamsInvité : Mike Hudack · Anime : Harry StebbingsThe Sling Money founder shares lessons from Facebook, Deliveroo and Monzo on customer choice, data and product leadership.
26:55ENA First Look at OpenAI o1-preview in September 2024Early tests, benchmark caveats and questions raised by the model’s system card.
26:28ENAPI prices fell 200x. The LLM table still is not set.Anime : Sarah Guo, Elad GilSarah Guo and Elad Gil, with no guest this week, argue over whether the language-model market has already consolidated — and why falling prices, five-person labs, and AMD’s shopping spree all cut against a simple ending.
1:13:44ZH孟醒:Waymo、Cruise 与无人驾驶的运营难题Invité : 孟醒 · Anime : 张小珺从凤凰城乘车体验到安全、车队和商业模式,回看 2024 年的行业观察。
21:21ENOpenAI o1’s 2024 Launch: Reasoning Benchmarks and Early TestsMatthew Berman · Forward FutureMatthew Berman reviews o1-preview and o1-mini, OpenAI’s reported evaluations, and his own early coding experiments.
3:42:21ENRome Always Raised Another ArmyInvité : Gregory Aldrete · Anime : Lex FridmanHistorian Gregory Aldrete tells Lex Fridman how a city of mud huts absorbed the Mediterranean, survived Hannibal, and then tore itself apart over the spoils.
Charger plus24 sur 4 441 affichés









