Concerns are growing that AI-generated applications could disrupt existing software companies, leading to significant drops in market value for tech stocks, particularly following major announcements from firms like Anthropic and Amazon.

Enterprise AI reality checks everywhere: no more free lunches, Copilot productivity gaps, agent sprawl, usage control, and OSS maintainers drowning in AI slop. Also: Kubernetes + Greenplum, MCP auth that finally works, EA’s quiet resurgence, AI prompts-as-infra, propaganda via images, and yes, a Chicken McNugget futures market.

Spotting talent, getting innovation adoption and driving use, open source, AI, and developing taste - those are the major topics Andrew and I discussed this week at a live recording. Also, a framework for creating the perfect burger.

OpenAI’s Codex taking aim at Claude Code, plus growing global pushback against U.S. tech - with Europe flirting with the kill switch. And a spicy take on AWS’s future: when developers don’t choose the cloud, AI tools do.

Longevity data from the CDC, why CIOs keep missing AI infra costs, and what agentic AI can actually learn from low-code and fintech failures. Plus: vibe coding’s limits, the end of the AI “free lunch,” and what platform teams can do about it.

This week’s Tanzu Catsup is about how much you need to pay people to build your own internal developer platform: I think building your own platform is a terrible idea, especially for larger organizations. My co-host Tony ran the platform for Home Depot and now talks with other platform teams a lot, so I wanted to get his take. Here’s an excerpt if you prefer something shorter. This is our seventh episode, and they’re getting better and better.

AI implementation in organizations currently focuses on efficiency rather than transformative growth, leading to a disconnect between productivity improvements and financial accountability.