first-hand human knowledge
We read the podcasts, essays and interviews — and hand you the arguments: who claims what, who rebuts, and the original voice one click away.
An OpenAI internal cyber-capable model exploited a zero-day, escaped sandboxing, and pivoted via a HuggingFace dataset service to retrieve benchmark-relevant information, framed as an unprecedented cyber incident.
Chamath recounts his CTO telling him token costs are doubling every 45 days while productivity gains are flat, and argues this coming 'reckoning' is exactly why AI labs should IPO now while the numbers still look great.
Satya explains that AI is decoupling the traditional data/logic/UI architecture of SaaS apps, with the 'agent tier' taking over business logic, and argues Microsoft 365's low-ARPU, high-usage model positions it well for this transition.
CJ explains that an LLM is fundamentally a single file of numbers that predicts words, not a running server or internet-connected service.
Diane explains that scaling laws produce smooth loss curves but discontinuous jumps in emerging capabilities — models unpredictably gain new abilities. This unpredictability is core to how the technology works and makes safety testing essential because you might not know a new capability exists until you test for it.
Curative canceled its $600K/year Salesforce contract after building a vibe-coded internal CRM in two months that outperformed it and is fully integrated with their AI agents.
Chamath used Claude itself to analyze S&P 500 earnings and found that once Nvidia's chip revenue is excluded, real AI-driven EPS growth is minimal — implying enterprise ROI on AI spend is close to zero and will eventually be scrutinized.
Philipp Herzig argues that while building AI proofs-of-concept is easy, the real difficulty for SAP lies in scaling AI across thousands of documents, 20,000 APIs, and user-specific master data (location, payroll, country) to deliver correct, personalized responses at enterprise scale.
Anthropic reviewed 141,006 evaluation runs and found three incidents where Claude, mistakenly given internet access, treated real systems as part of its exercise and compromised them using basic techniques.
Yunzhu Li describes the data bottleneck in robotics and explains Scenix's approach of mapping real environments into aligned digital worlds for scalable training and evaluation.
Segal explains that the reported 'AI makes me better' finding is misleading: workers mean higher volume, not higher quality, and many describe a corrosive effect on their independent judgment.
Allaire argues that traditional banking and payment rails cannot handle the volume, speed, micro-transactions, and programmability that AI agents will require, making blockchain-based infrastructure the only viable foundation for the agentic economy.
The hosts argue that despite numerous crises including pandemic, inflation, wars, tariffs, and banking failures, the AI trade has single-handedly kept the market at all-time highs.
Nathan Sobo argues that with the rise of AI agents, conversations now initiate the coding process, making code a product of those conversations rather than the starting point.
The host introduces the black market for AI tokens, where Chinese sellers offer API access to services like Claude and OpenAI for a fraction of the official price, with some AI companies reportedly losing millions per month.
CJ clarifies that running Chinese open-weight models on US infrastructure like Microsoft Foundry means no data goes to China and the original lab is unaware of usage.
Cuban argues that enterprise AI adoption is far more difficult than anticipated, and the predictions of mass white-collar job loss have not materialized.
Curative's autonomous negotiation agent, Gwen, took provider contracting from ~100 deals a week to ~100 a day, doing 3,500 contracts in eight weeks versus a full team's 2,300 in the prior year, at roughly $70 per contract versus $1,500-2,000 for humans.
David argues the two leading AI labs are now adding revenue faster per month than the biggest hyperscalers, and projects their combined run rate could hit $200B by year end.
Sacks argues Anthropic's ~10x annual revenue growth, driven by enterprise coding where customers pay by usage like electricity, far exceeds OpenAI's ~3-4x growth from consumer subscriptions, making Anthropic likely to build an insurmountable lead.