New here?
Start with these five
Chain of Thought is a show for engineers, founders, and operators shipping AI in production. These five episodes cover the threads that run through everything else: agents, evals, memory, infrastructure, and what it takes to ship a product. Read the transcript or press play.
- 5 episodes to start
- 4 hr 7 min of listening
- 77 in the archive
-
New to the terminology? Read AI vocabulary in plain English: 14 terms with business examples before you press play.
-
Prefer a route for your work? Choose a learning track by role, from business leadership to production engineering.
-
You can also read a shorter explanation first. Types of AI agent memory accompanies the memory conversation, and the evaluation hub collects more guests' approaches to testing. Use the glossary when a term is unfamiliar.
The five episodes
Choose the conversation closest to the decision in front of you. Jerry Liu discusses the context layer around agents; Alex Ratner examines evaluation; Richmond Alake focuses on memory. Paige Bailey takes the discussion inside model and tooling experiments, while Charity Majors brings it back to operating software and solving business problems.
Each episode's transcript lets you return to the original discussion rather than relying on a summary.
-
The AI Framework Era Is Over: Why Context Is the Moat | Jerry Liu of LlamaIndex Jerry Liu, LlamaIndex LlamaIndex founder Jerry Liu on why the advantage in agents moved from the framework to context. The cleanest entry into how we think about agents. -
Every AI Agent Has an Evaluation Gap | Alex Ratner, Snorkel AI Alex Ratner, Snorkel AI Snorkel AI’s Alex Ratner on how agent capabilities have outpaced evaluation, and the benchmarks his team is building to measure them. -
Agent Memory: The Last Battleground in the AI Stack | Richmond Alake, Oracle Richmond Alake, Oracle Richmond Alake on engineering memory for agents that need to carry information across sessions, including how they should forget. -
Gemini 3 & Robot Dogs: Inside Google DeepMind's AI Experiments | Paige Bailey Paige Bailey, Google DeepMind Google DeepMind’s Paige Bailey on what a frontier lab is shipping: models, tooling, and the experiments behind them. -
AI Won't Solve Your Toughest Engineering Problems | Honeycomb’s Charity Majors Charity Majors, Honeycomb Honeycomb’s Charity Majors on production reliability and why AI won’t solve your hardest engineering problems.
A Chain of Thought experiment
Try an agent review
An AI agent proposes a bug fix and changes the failing test. Inspect two pieces of evidence, make a release decision, and compare the costs in a fictional incident.
Latest episode · EP 77
Switching Models Shouldn't Mean Starting Over | Walrus Protocol's Kimberly Logan
Kimberly Logan, Walrus Protocol ·