YoutuBlog (HackerTube) : lire les conférences et podcasts de YouTube
Derniers articles
6 135 articlesTrier
15:21EN
AI Engineer· 9 nov. 2026Ashwin Paranjape: Testing financial agents against earnings consensusAshwin Paranjape · Samaya AISamaya compares seven models on more than 450 earnings releases using time-limited data, a bias-corrected baseline and several scoring methods.
17:57ENAI Engineer· 9 nov. 2026Maty Bohacek: Can AI agents impersonate people to pass verification?Maty Bohacek · Stanford UniversityA controlled vault experiment tests whether agents use available audio and video impersonation tools, and how safety alignment changes the result.
50:17EN
20VC with Harry Stebbings· 9 nov. 2026Podcast · 2 voixMistral’s Arthur Mensch on sovereign AI and the compute bottleneckInvité : Arthur Mensch · Anime : Harry StebbingsThe CEO says customers need model choice and continuity, while European infrastructure must scale ahead of demand.
18:56ENAI Engineer· 9 nov. 2026Pratyusha Duvvuri: Making AI token spending understandable to engineering and financePratyusha Duvvuri · CiscoA Cisco engineering lead explains why provider bills, team ownership and return on investment require different views of the same AI usage.
45:21EN
Theo - t3․gg· 9 nov. 2026Theo: AI coding makes maintaining the development environment the jobTheo Browne · Ping LabsFrom Twitch to T3 Code, Theo argues that engineers must remove the friction and feedback failures that block agents.
16:57EN
Every· 9 nov. 2026PodcastFable 5.2 excels at ambitious builds, but its token costs add upEvery’s early tests found strong coding and creative work; a single documentary run used 1.4 billion tokens.
19:24ENAI Engineer· 9 nov. 2026Richard Feldman: keep the agent conversation with the codeRichard Feldman · DeltaA shared thread can preserve decisions, reviews and handoff context that otherwise gets lost between teammates.
15:28ENAI Engineer· 9 nov. 2026Ryan Carson: One manager agent coordinates 20 coding agentsRyan Carson · UntangleUntangle delegates coding, review and merge coordination to agents, but its founder watches each feature before approving shipment.
1:50:34ENAI Engineer· 9 nov. 2026A financial AI agent must show its workFrancesco Galletta, Felippe Felisola Caso · MonzoA wedding-budget demo pairs tested calculations with evidence, decision gates and independent checks.
19:50ENAI Engineer· 9 nov. 2026Keiji Kanazawa rebuilt his agent system after reading its codeKeiji Kanazawa · MicrosoftA three-week refactor produced fake agents and expensive tests. He rebuilt around explicit goals and evaluations.
20:51ENAI Engineer· 9 nov. 2026Jason Mancuso: Scaling agent reinforcement learning across mixed GPUsJason Mancuso · ModalModal combines asynchronous rollouts, mismatch corrections and compressed weight updates to train coding agents across a fragmented GPU fleet.
24:47EN
How I AI· 9 nov. 2026PodcastClaire Vo: Fable 5.2 excels at creative builds, but not every designA marketing animation, furniture guide and 3D typewriter impressed Vo; independently generated website designs did not.
24:18EN
Matthew Berman· 9 nov. 2026Fable 5.2 built playable demos—but several cost thousandsMatthew Berman · Forward FutureMatthew Berman shows games, creative software and simulations, with estimated API costs from $25 to $15,000.
18:32ENAI Engineer· 8 nov. 2026Why a plausible contract analysis missed $800,000Natalia Connolly, Kevin Madura · AlixPartnersAlixPartners’ Natalia Connolly and Kevin Madura show how missing links, renewal clauses and silent access errors can derail an AI procurement analysis.
23:15EN
Latent Space· 8 nov. 2026Podcast · 2 voixJay Parikh: GitHub must adapt as coding agents change developmentInvité : Jay Parikh · Anime : Shawn WangGitHub's scale, model choices and security are changing as agents join developers, says Microsoft CoreAI's Jay Parikh.
58:33EN
Lenny's Podcast· 8 nov. 2026Podcast · 2 voixGemini’s product chief on testing models and building ahead of themInvité : Tulsee Doshi · Anime : Lenny RachitskyTulsee Doshi explains a small-eval starting point, her 60-20-20 product framework and why model behavior itself is a user experience.
46:40EN
Peter Yang· 8 nov. 2026Podcast · 2 voixAshley Alexander: How AI Helps Me Care for My Parents With CancerInvité : Ashley Alexander · Anime : Peter YangA caregiver uses AI to prepare for appointments, organize records and ask better questions—while keeping doctors at the center of treatment.
12:38EN
Chris Williamson· 8 nov. 2026Podcast · 2 voixWhat makes AI-generated work feel like slop?Anime : Chris WilliamsonChris Williamson and his guest ask creators to explain their judgments, and to distinguish a tool’s defaults from intentional craft.
18:56ENAI Engineer· 8 nov. 2026How Citadel Securities builds a memory that catches troubled projectsJoao Wenzel · Citadel SecuritiesJoao Wenzel describes an agent system that weighs evidence, tracks aging facts and tests its warnings against what happened next.
18:39ENAI Engineer· 8 nov. 2026How Ramp tests a bookkeeping agent before and after launchRyan Stevens · RampRyan Stevens explains synthetic businesses, all-or-nothing scoring, skill ablation and a two-period test for agent memory.
16:09ENAI Engineer· 8 nov. 2026How Range’s financial planners built their own AI workflowsMatt Ragone · RangeMatt Ragone’s team used real cases and existing advice standards to build skills, while keeping a certified planner’s review before client delivery.
20:28ENAI Engineer· 8 nov. 2026Why AI coding gains depend on more than faster code generationMartin Harrysson, Prakhar Dixit · McKinseyMcKinsey speakers describe smaller teams, shorter loops, curated context and quality checks observed among enterprise clients.
1:21:42EN
Y Combinator· 8 nov. 2026What AI biology can do now, from cancer data to DNA designAlfredo Gonzalez, Yusuf Roohani, Garyk Brixi, Divya NoriFour researchers explain patient-specific cancer analysis, virtual cells and genome models, separating early experiments from proposed clinical applications.
31:22ENAI Engineer· 7 nov. 2026Nouha Dziri on Making Coding-Agent Teams More Cost-EfficientNouha Dziri · CohereExperiments on ProgramBench and Commit0 show why the task bottleneck, lead model and coordination design matter more than simply adding agents.Charger plus24 sur 6 135 affichés