Summary In today's episode, I tackle Cecil's question about the best strategies for managing LLM context window decay and token limits during long-running analytical sessions. Here's what this means for you. You'll discover that offloading work to deterministic tools and persistent notes lets you squeeze dramatically more value out of any LLM you use. You'll…
_Almost Timely News: 🗞️ How To Improve Advertising with AI (2026-06-28) :: View in Browser_ The Big Plug 👉 My new course, GEO 201 on competitive GEO measurement, is now for sale. Content Authenticity Statement 100% of this week’s newsletter was made by me, the human. In the video version, you’ll see Claude outputs. Learn…
SpaceX/xAI has acquired Cursor. What does that mean for your data privacy? If you are NOT an enterprise customer, and you are using private data with Cursor, now is the time to find a competing product and request deletion of your data, before the deal fully closes. DISCLAIMER: I AM NOT A LAWYER. I cannot…
Summary In today's episode, I walk through the top five strategic use cases of job descriptions beyond just listing job duties, using McKinsey as a real-world example. Here's what this means for you. You'll discover how a single batch of postings can reveal a company's tech stack, strategic direction, and even your own job's vulnerability…
Is the AI bubble going to burst? Depends on what you mean. Is the absurd hype, the crazy valuations, and the insane speculation around AI going to burst? Without a doubt, yes. The trillions of dollars flowing into things like data centers and software companies are one math paper away from complete obsolescence, something I’ve…
Summary In today's episode, I explore whether offloading cognitive tasks to AI in the enterprise is creating dead workplaces and what you can do to avoid that trap. Here's what this means for you. You can defend your executive function by handing templated work to machines and doubling down on non-templated, judgment-heavy tasks. You'll also…
Summary In today's episode, I answer whether replacing mid-career staff with AI risks losing institutional knowledge and explain why your corporate knowledge bases always miss the failure history that actually lives in your employees' heads. Here's what this means for you. You discover that AI systems inherit hidden survivor bias when you train them on…
The appliances keep getting better. So why isn’t our food? This is the situation a lot of folks are experiencing right now with AI. OpenAI released a new version of Codex today that tries to be Claude Cowork. OpenCode has a slick new UI. Claude Code has new dynamic workflows, in the style of Paperclip…
Summary In today's episode, I explain why enterprises need both contractual agreements and technical safeguards to protect intellectual property from AI model training. Here's what this means for you. You gain a layered defense strategy that closes off the most obvious vulnerabilities before they turn into data leaks. You'll also learn these concepts: how the…
