Artículos recientes
4131 artículosOrdenar
1:18:07EN20VC Roundtable: AI Spending, Model Competition and Software AcquisitionsInvitado: Jason Lemkin, Rory O’Driscoll · Presenta: Harry StebbingsHarry Stebbings, Jason Lemkin and Rory O’Driscoll debate enterprise AI returns, prediction markets, SaaS roll-ups and the cost of raising capital.
1:01:27ENTwo Years and Four Months to a Live Reactor: Valar Atomics Wants Energy Ten Times CheaperPresenta: Sarah Guo, Elad GilIsaiah Taylor walks Sarah Guo through Ward 250 in Utah and argues that nuclear is a hardware-iteration problem, not a paper-design one.
1:29:02ENFrom 50 Designs a Day to 1,500: How AI Changed What Jaguar Land Rover Can ExplorePresenta: Nathan LabenzNeural Concept’s Thomas von Tschammer on specialist physics models, engineering copilots, and why Formula 1 already runs thousands of designs overnight.
28:37ENTheo Tests Sonnet 5 and Explains Fable’s ReturnTheoTheo compares model benchmarks and his own coding tests, finding promise in Sonnet’s subagent orchestration but questioning its task costs.
2:19ENTheo on Sonnet 5: Agent Orchestration and Token CostsTheoTheo welcomes Sonnet 5's ability to coordinate subagents while questioning its token efficiency. He also discusses the introductory pricing announced at release.
31:30ENWhy Some AI Models Use Fewer Tokens to Solve a TaskTheoTheo explains input, output, caching and reasoning tokens, then explores a hypothesis about terse internal reasoning.
0:33ENTheo on Shared AI Conversations and Team CollaborationTheoA shared Claude conversation in Slack lets teammates see and continue the same work. Theo argues that this kind of collaboration differs from a GitHub-centered workflow.
3:51:47ENForty-Six Percent of Emperors Were Overthrown. The Empire Lasted AnywayInvitado: Anthony Kaldellis · Presenta: Lex FridmanHistorian Anthony Kaldellis walks Lex Fridman through 2,200 years of Roman history and argues that what we call Byzantium was Rome itself.
1:10:15ENThe 100x lives between the layers: Dylan Patel on co-design and a 60x yearSemiAnalysis’s founder argues that matching models, kernels, and silicon is what turns a 2x here and a 2x there into 100x, while equivalent-quality inference has gotten about 60 times cheaper in a year.
0:42ENMatthew Berman's Concern About Access to Frontier AIMatthew Berman · Forward FutureIn a short commentary, Berman argues that restricted model access could disadvantage startups and slow US innovation.
1:06:10ENK. R. Sridhar on Distributed Power, AI Demand and Energy AccessInvitado: K. R. Sridhar · Presenta: Harry StebbingsBloom Energy’s CEO discusses data-center power, manufacturing, natural gas and his vision for broader electricity access.
22:43ENMac Prices, Unified Memory and Local AI: Theo’s Buying Trade-offsTheoTheo explains why high-memory Macs appealed for local inference and how rising component prices changed his comparison with PCs.
12:26ENHermes Agent: Hosted Setup, Skills, Scheduled Tasks and TelegramMatthew Berman · Forward FutureIn a sponsored June 2026 tutorial, Matthew Berman demonstrates a managed Hermes installation and explores its tools, memory and messaging integrations.
30:08ENGPT-5.6’s Limited Preview: Benchmarks, Agent Risks and AccessTheoTheo reviews the launch-era announcement, published evaluations and reported agent incidents, while questioning the balance between safety and broad access.
0:25ENApple’s June 2026 Price Increases and the Memory ShortageMatthew Berman · Forward FutureMatthew Berman links higher Mac and iPad prices to the demand for memory driven by AI infrastructure.
54:29ENfomo CEO Paul Erlanger on Social Trading, Fundraising and Product GrowthInvitado: Paul Erlanger · Presenta: Harry StebbingsErlanger discusses building a trading app, raising from angels and venture firms, and combining social features with financial products.
1:57:46EN30% Consciousness, 47% Bee, 80% Turnover: Putting Numbers on What AI Is BecomingPresenta: Nathan LabenzA Cognitive Revolution weekly cut that starts inside the model with Cameron Berg, then zooms out through gradual disempowerment, Europe’s bind, and the companies rebuilding around agents.
1:08ENMatthew Berman Criticizes GPT-5.6’s Restricted PreviewMatthew Berman · Forward FutureA June 2026 reaction to the government-requested staged release argues that limited access disadvantages independent builders.
19:49ENMatthew Berman on Restricted AI Model Releases and the Risk of Concentrated PowerMatthew Berman · Forward FutureA June 2026 commentary argues that limited previews for frontier models could disadvantage startups and independent developers.
0:48ENTheo Responds to Karpathy’s Discussion of Claude in SlackTheoTheo says a post about tagging Claude in Slack may seem surprising, but he sees lessons for the AI industry in the feature.
1:02ENAnthropic Accuses Alibaba-Linked Operators of Distilling ClaudeMatthew Berman · Forward FutureMatthew Berman discusses a 2026 allegation, restrictions on advanced models and his prediction about identity checks.
36:19ENThe Benchmark Grid Is Broken: Why Test-Time Compute Has Outrun How We Score ModelsPresenta: Sarah Guo, Elad GilOpenAI researcher Noam Brown argues that static leaderboards hide how long a model is allowed to think — and that safety evals, research taste, and the release cycle have not caught up.
19:24ENTheo on Google’s Coding Agents, Departures and Internal ExperimentsTheoTheo distinguishes model knowledge from agent behavior and argues that organizations need room for useful prototypes.
20:47ENFrom Chatbots to Team Agents: Theo on Karpathy’s Claude Tag ThesisTheoTheo explores shared channel context, asynchronous agents and the trade-off between a ready-made team workflow and model choice.
Cargar másMostrando 24 de 4131









