Why do AI analytics tools look impressive… but still don’t get used?
James Carman talks with Pete Hunt, CEO of Dagster Labs and one of the original engineers behind React, about what it really takes for AI-powered analytics to stick. They unpack why adoption is often the biggest challenge (not accuracy), how bringing analytics into Slack changes decision-making, and what “fast, directionally correct” can unlock when teams trust the path to an answer.

If AI can build a prototype in a day, why does shipping still take months?
In this episode, James Carman is joined by David Arenas, a senior technology leader across industrial automation, fintech, and global data platforms. Together, they unpack why AI is accelerating development, but also making the gap between demo and production more expensive. They break down what it takes to ship AI-built work at enterprise scale.

What did we actually learn about AI this year?
To close out 2025, James interviews an unexpected guest: ChatGPT. They talk honestly about what teams got wrong, what actually improved, and why AI’s biggest wins came from small workflow shifts — not sweeping transformation. They also dig into software engineering, documentation, onboarding, junior engineers, and what tech leaders will need to operationalize in 2026 as AI becomes a system, not a tool.