Free — sign in with your email, no password.
Podcasts on 10 November 2026
Meredith Whittaker argues consumer AI agents' root access and prompt injection make them untrustworthy, while Neil Chilson says conditional, tiered use and better-built agents can improve privacy over the status quo.
8 key points · 1 quoteAWS CEO Matt Garman argues AI is pulling on-prem workloads to the cloud, fueling a path toward $1 trillion in revenue, while AWS stages capex carefully and bets on layered offerings and provable agent safety.
10 key points · 2 quotesStephen Witt argues Jensen Huang's affair with an alleged Chinese spy, tied to chip smuggling, exposes Nvidia's key-man risk, erodes his credibility on China, and weakens the AI-optimist camp.
8 key pointsA Monday news show covering Jensen Huang's admitted affair with a suspected Chinese spy and its fallout for Nvidia and chip policy, plus guests from Sweetgreen, Nominal, Foundry Robotics and Coframe.
8 key points · 1 quoteJensen Huang's admitted affair with a suspected Chinese spy is a geopolitical and market story that threatens Nvidia's China lobbying, even as AI funding valuations soften and hyperscalers favor powered, ready data centers.
6 key pointsStripe argues SaaS platforms are growing despite the 'SaaS apocalypse' and can keep winning by optimizing payments, becoming customers' financial home, and monetizing AI with usage-based pricing.
8 key pointsSam argues AI makes software cheap to build, so durable value comes from taking on customers' risk and running whole businesses from the P&L backwards, with tiny teams and custom software.
8 key pointsSean argues that software correctness should be built in from the start through strong type systems, models and workload-driven testing, not checked at late quality gates, and that pragmatism about debt and speed matters.
10 key points · 1 quoteBrendan Foody argues that as agents automate repetitive work, humans will contribute by teaching models through evals, rubrics and RL environments, making expert data a durable, growing market that Mercor aims to lead.
11 key points · 2 quotesDharmesh Shah argues that startups can win outside the frontier labs' path by solving whole customer problems in narrow niches, treating AI as infrastructure, and learning cheaply through fast iteration.
9 key points · 1 quoteTubi and e.l.f. beat giants by serving overlooked consumers, refusing to play incumbents' game, building a new one in the white space, and staying close to their communities.
7 key pointsEvery's team rates Fable 5.2 a gold-medal, paradigm-shifting model for long-running, ambiguous, high-judgment builds, while Opus 5.5 stays the daily driver and the model stumbles on tightly specified, simple tasks.
12 key pointsFormer Tesla president Jon McNeill explains how speed, first-principles simplification, blunt communication and customer closeness let Tesla grow about 10x with no marketing budget, and how he applies those lessons elsewhere.
12 key points · 3 quotesShane, worth about $10M at 57, argues that freedom and time matter more than extra zeros, and that health, happiness and example-setting are the real measures of wealth.
12 key pointsThe hosts argue Jensen Huang's admitted affair with an alleged Chinese spy is a credibility-destroying blunder that weakens his China chip advocacy, though his policy stance long predated it.
5 key pointsCody Schneider argues that marketing is now code and tokens, so giving an AI agent data, platform access and tools lets it produce static and UGC ads at scale, cheaply.
9 key pointsOpus 5.5 and related models have done for video editing what Opus 4.5 did for coding, making agentic video viable for ordinary knowledge workers, who should try it on internal and external communication.
11 key points · 4 quotesClaire went from disliking Claude models to loving Fable 5.2, a fast, creative, non-annoying frontier model strong at animation, geometry, long-running coding and one-shot apps, but weak at web design.
9 key points