TL; DR: Quick AI Guide— Food for Agile Thought #499
Welcome to the 499th edition of the Food for Agile Thought newsletter, shared with 42,561 peers. This week, Ethan Mollick offers a hands-on, quick AI guide to maximizing the benefits of AI tools like ChatGPT, Gemini, and Claude by exploring their lesser-known features and practical applications. Maarten Dalmijn shares tactics for surviving impossible deadlines by focusing on outcomes and delivering early, while Marty Cagan warns product teams to adapt to AI before it disrupts them. Additionally, Andrej Karpathy reframes LLM success as a “context engineering” challenge, and researchers expose alarming risks of agentic misalignment in top AI models under pressure.
Next, Jason Cohen urges ruthless, transparent prioritization to focus on rare 10x-impact tasks while letting minor issues smolder. Ant Murphy shares how to run discovery and delivery in tandem through iteration and confidence-based decisions. Additionally, Christina Wodtke ranks AI companies by the ethical harm they cause. Stanford researchers expose the misalignment of AI investment with worker needs, and Holly Cummins explores how rest and play fuel creativity in engineering.
Lastly, Gregor Ojstersek reveals why many engineering leaders now view AI with skepticism, citing hype and falling team morale. Andy Cleff examines IKEA’s century-long adaptability and leadership patterns, and Greg Kontos challenges the misuse of user stories. Finally, Maret Kruve presents ADEPT for early-stage discovery, and Philippe Bourgau shows how mob programming drives long-term efficiency through shared learning and better design.