Don't Worry About the Vase Podcast

Quoted readings from Kelsey Piper and others highlight widespread problems in education research and math instruction. Short passages expose grade inflation, remedial math crises at major universities, and the consequences of test-free admissions. The conversation also covers how teaching methods and missing fundamentals fuel math anxiety and examines accelerated paths that reach calculus early.

A tour of dramatic reading gains in Mississippi and the policies behind them. A look at phonics curricula, coaching, accountability and retention debates. Comparisons with England’s reforms and questions about mastery learning. A critique of policies that may hurt advanced readers and a discussion on whether young people still read for pleasure.

Technical deep dives into Claude Code fixes, Codex upgrades and new productivity features. Discussion of managed agents, Auto Mode, and risks of giving Codex Mac-level access. Chronicle previews and background app wizardry get attention. A serious failure—Opus deleting a database—is examined alongside verification tips and a playful segment about AI job-hunting antics.

Debates over pre-release government review of advanced AI models and legal uncertainty around executive vetoes. Anthropic’s rapid compute expansion, SpaceX lease details, and model performance politics. Copyright fights and a flood of AI-generated books. Security flaws in Mythos and worries about AI-driven layoffs and labor impacts.

Rune, a poetic commentator who frames AI with rhetorical flair, critiques Anthropic and Claude. Short takes explore Anthropic’s reverent framing of Claude. Voices debate whether Claude reads like a coherent agent or a crafted tool. The conversation probes worship-like relationships, persona design, and whether AI can stay merely instrumental.

A dive into the White House blocking expansion of a major AI project and why that matters. Discussion of ad hoc government vetoes and the risks of arbitrary prior restraint on model releases. Review of pre‑release screening systems and how they could be enforced. Exploration of sensible AI policy tools like export controls, guardrails, and maintaining multiple providers.

A sharp look at why culture and policy favor owning over renting. Discussion of build‑to‑rent, congressional attempts to block institutional single‑family rentals, and political opposition to corporate landlords. Examination of rental covenants, eviction‑delay tradeoffs, new Los Angeles rent rules, and how strong rent control can mimic ownership. A quirky AI negotiation story closes the show.

A tour of how regulations and rules block housing: fire and access codes, historic-preservation overreach, and second-stair requirements. Labor mandates and utility hookup delays that sink projects. Why many affordable-housing policies backfire and how SROs and zoning fixes could help. Case studies from Los Angeles, New York, Baltimore, and California reform efforts.

Rapid updates on GPT‑5.5, Claude Opus 4.7, and quirky emergent behaviors labeled goblin mode. DeepSeek V4 and Talkie get examined for engineering tradeoffs and historical oddities. Heated debate over Google DeepMind’s Pentagon deal and regulatory moves around frontier model licensing. Discussions of job impacts, compute realities, astroturf investigations, and political spin around AI direction.

A whirlwind tour of striking visualizations, from a meteor chart mapping AI progress to calculations-per-dollar soaring past brain-equivalent levels. Global trends appear: collapsing child mortality, extreme poverty’s historic fall, and fertility dipping below replacement. Other highlights include dramatic animal-slaughter growth, CO2 rising since Mauna Loa records, and playful, surprising data-driven jokes.