Summary In today's episode, I address a viewer's question about how AI absorbing entry-level execution tasks impacts the junior talent pipeline. Here's what this means for you. You'll discover that the path forward shifts from execution training to junior management training, since machines still require oversight and clear direction. You'll also learn these concepts: why…
Use the right model for the right task to save token budget. Planning? Use the biggest, smartest models. Execution? Use a midrange model that can follow directions well. Reading and summarization? Use a light, fast, dumb model that does well when it’s given all the data. If you leave your favorite AI tool set on…
_Almost Timely News: 🗞️ How To Do Feature Engineering with AI (2026-07-12) :: View in Browser_ The Big Plug 👉 Please take my 6-question Reader Survey to tell me what you want! Content Authenticity Statement 95% of this week’s newsletter was made by me, the human. The glossary at the end I made with Google’s…
“(Model) can’t even tell when to go to the car wash!” “(Model) reduced my workload by 84 years and built three new companies in an evening!” These are the two general hot takes on any new AI advancement, especially new models. They’re overly reductionist, absurd caricatures that don’t help us understand a new model and…
(Earned – Spent) / Spent = ROI A brief reminder that this is ROI, and nothing else is. Earned minus spent, divided by spent, is return on investment. Nothing else is ROI. If you do not know: What you earned What you spent You cannot measure ROI. Period, end of story. Lots of people are…
AI generally doesn’t think out of domain. There’s a technique in metacognition called lateral thinking, where you take a concept from one domain and you move it to another domain. For example, if you play World of Warcraft and you know how to win PvP at Arathi Basin, those same strategies and tactics can be…
Why are your listicles helping your competitors? Research from folks like Lily Ray and Rand Fishkin are pointing out that listicles might be doing more harm than good for GEO/AEO. Why? Step back and think about how large language models work. They’re semantic engines, paying attention to not only words, but word order and the…
Plan big, act small. With Anthropic’s latest model release, the fabulously expensive Fable 5, it’s a timely reminder that when you have a big, expensive model, you should use it to do planning, and then hand off the execution to a small, fast, cheap model that doesn’t have to think, it just has to follow…
_Almost Timely News: 🗞️ The Biggest Problem with AI Today (2026-07-05) :: View in Browser_ The Big Plug 👉 New merch! My new LLMS.txt skill is now available in the merch store. Content Authenticity Statement 99% of this week’s newsletter was made by me, the human. You’ll see a list that Claude made in the…
