🇬🇧 EN
Complete
Jensen Huang describes AI as moving from generating or translating information to reasoning and doing work, arguing that this shift requires computers that generate intelligence in real time. · He presents Nvidia’s AI systems as factories that turn electricity into tokens, and outlines five investm…
Read summary
🇬🇧 EN
Complete
🇬🇧 EN
Complete
The guest presents SpaceX’s launch capability, especially Starship’s planned rapid reusability, as the foundation for expanding Starlink, direct-to-cell service, and AI compute in space. · The interview’s investment thesis centers on claimed cost and product advantages, existing connectivity demand…
Read summary
🇬🇧 EN
Complete
The speaker views the sharp weekly decline as a needed rotation after a crowded rally, not the start of a bear market, while expecting a choppier market that rewards selectivity. · They favor AI applications and infrastructure with longer runway, especially Eli Lilly and Marvell, have sold Micron,…
Read summary
🇬🇧 EN
Complete
Lynn Alden describes G7 debt and bond-market pressures as a slow-moving fiscal problem, while warning that a prolonged Strait closure could exhaust reserves and cause higher energy prices, shortages and food-price pressure. · She sees real demand behind parts of the AI boom but expects growth and r…
Read summary
🇬🇧 EN
Complete
Demis Hassabis describes AI as accelerating science through fast solutions, broad access to results, and potentially faster discovery, while warning that engineering progress is outpacing scientific understanding. · The panel discusses where AI can help research, how it might free scientists for cr…
Read summary
🇬🇧 EN
Complete
Visser says Bitcoin remains a long-term portfolio holding but is in a bear market after falling 50% from its high; he would look for a base and a move above long-term moving averages before adding conviction, while continuing to buy in small amounts and favoring diversified exposure. · He sees AI a…
Read summary
🇬🇧 EN
Complete
The speaker presents a token burn dashboard as a way to understand how AI use changes over time and to connect usage patterns with tasks and outcomes. · He describes building it with Codex, approximating Claude usage from artifacts because Claude does not readily expose token counts outside the API…
Read summary
🇬🇧 EN
Complete
Matt Cole argues that Bitcoin-backed digital credit can offer high yield with less volatility than Bitcoin and may help investors move from fiat toward a Bitcoin-based future. · Strive’s approach uses perpetual preferred equity, reserves, and adjustable dividend payments, while accepting risks incl…
Read summary
🇬🇧 EN
Complete
Bob Lucas argues Bitcoin is following a typical four-year bear-market cycle: after a countertrend rebound to just under $83,000, it has fallen about 25% and retested the February lows near $60,000. · The model portfolio bought 10 BTC at $65,000, bringing its allocation to 58% BTC and 41% cash; the…
🇬🇧 EN
Complete
The speaker argues that Opus 4.8 is a strong model with gains on long-running tasks, but its reasoning can be inconsistent and its performance depends heavily on the product harness around it. · From their tests and use, they prefer GPT-5.5 in Codex for complex, long-running work, while identifying…
Read summary
🇬🇧 EN
Complete
Andrew Kang argues that advances in physical AI could make adaptable robots economically viable within a few years, opening a market tied to human labor and potentially supporting very large robotics businesses. · He backs that thesis with investments in humanoid and other robotics companies, while…
Read summary
🇬🇧 EN
Complete
At a Stanford conversation, Demis Hassabis described how experiences in games, chess, neuroscience and engineering shaped his long-running goal of building AI as a tool for science. · He discussed AlphaGo and AlphaFold, including why AlphaFold’s protein structure predictions were made broadly avail…
Read summary
🇬🇧 EN
Complete
The conversation moves from Blockmates’ approach to media and its early coverage of Hyperliquid into debates about crypto launches, privacy, Ethereum, and Near. · Grant argues that crypto markets are responding to product quality, fairer access, privacy concerns, and the prospect of agent-driven pa…
Read summary
🇬🇧 EN
Complete
The speaker sees a strong but increasingly narrow market rally, with AI infrastructure driving gains while supply bottlenecks and concentration make the outlook more fragile. · They favor a thematic portfolio tied to AI infrastructure and related beneficiaries, have sold Micron and rotated into sil…
Read summary
🇬🇧 EN
Complete
The speaker argues that AI makes polished work easier to produce, weakening the value of portfolios and other finished artifacts as evidence of skill. · They recommend showing judgment through live, challengeable reasoning about a real problem, organized around situation, decision, risk, and change…
Read summary
🇬🇧 EN
Complete
Benedict Evans sees AI as a major platform shift on the scale of the internet or mobile, but says its uses, effects on jobs, and economic winners remain uncertain. · He argues that task automation does not by itself predict what happens to a whole job, and that value may accrue to applications and…
Read summary
🇬🇧 EN
Complete
Jensen Huang describes AI as a broad industry spanning energy, chips, infrastructure, models and applications, with agentic systems able to plan, use tools and act autonomously. · He argues that people should learn to use AI, predicting that it will change jobs and create new work, while calling fo…
Read summary
🇬🇧 EN
Complete
The speaker argues that AI makes software generation cheaper, shifting product management from rationing engineering work to judging, classifying, and governing the resulting abundance. · PMs should understand AI systems and markets, welcome broad experimentation, and decide which software to suppo…
Read summary
🇬🇧 EN
Complete
The speaker argues that agent products need product analytics centered on delegated work and agent runs, beyond chat activity and engineering traces. · They recommend tracking run starts, task completion, and user corrections under a shared run ID, while distinguishing completion from user acceptan…
Read summary