Latest · ai The Apocalypse Has a Permitting Problem A famous forecast says AI might end the human era within a few years. I went looking for the ending where the rest of us get a say, and couldn't find it. read the post → Earlier · ai AI Adoption Needs a Transition Plan A six-pillar framework for adopting AI while expanding human capability, improving work, and sharing the benefits of technological change. read the post → Earlier · ideas Still Aimed at Humans I've written the version of the future that worries me. This is the part I'm actually hopeful about: what happens to work, and why the human doesn't go away. read the post → Earlier · everything-else What I Love About Calvin and Hobbes The last strip hangs framed on my office wall. I've been a fan my whole life, and it's the rare thing from childhood that got better as I got older, not worse. read the post → Earlier · ideas Intelligence-First You think you're a digital company. That was the last war. The next one goes to whoever builds intelligence into how they actually operate. read the post → Earlier · ideas I've Thought Hard About Where AI Leads, and None of It Looks Easy I build with this technology for a living, which is exactly why I can't sell you a clean utopia or a tidy apocalypse. The honest map has no obviously good roads on it. read the post → Earlier · ideas A Better Mental Model for Understanding AI Stop thinking of it as a person or a database. Think of it as a very fast, very literal intern who has read everything and only knows what you put in front of it. read the post → Earlier · work Running Off the Inbox, Not a Plan Reactivity feels like productivity. It rarely is. read the post → Earlier · ai Why Most AI Pilots Fail It's almost never the model. It's the plumbing and the goal. read the post → Earlier · ops The Meeting That Keeps a Company Honest The meeting that keeps a company honest, step by step. read the post → Earlier · ai Good Enough Runs Local Now The assumption that serious AI means renting a frontier model by the token is already out of date. For most of what businesses actually need, the model can live in your own building. read the post → Earlier · work Positioning Is a Verb, Not a Noun Most founders think positioning is the reputation they'll earn someday. It's actually a decision you make on Monday: who you serve, and how you win. read the post → Earlier · ai Who's Liable When the Machine Is Confidently Wrong? These systems don't fail like software fails. They fail by being fluent, confident, and wrong, and we haven't decided who owns that yet. read the post → Earlier · ai The AI Bubble Is Real. So Is the Technology. A bubble and a genuine technological shift are not mutually exclusive. Both things are happening at once, and the people who understand that will still be standing when the froth clears. read the post → Earlier · ops Selling to Slow, Careful, Regulated Buyers The buyers who move slowest are also the hardest to dislodge once you're in. Here's how to get in, and why their caution becomes your moat. read the post → Earlier · ideas Automate the Drudgery, Keep the People There are two ways to sell AI to a business. One counts how many people you can fire. The other counts how good your people can become. I only do the second one. read the post → Earlier · dev Edge Cases Are the Whole Job The demo always works. The demo is easy. Everything interesting, and everything that pays, lives in the weird 1% where a white dog gets clocked at 90 miles an hour. read the post → Earlier · ai Why Most Enterprise AI Projects Quietly Fail The failure rate for corporate AI programs is staggering, and almost none of it is about the technology. It's about the people you forgot to bring with you. read the post → Earlier · ops Show Me Your Day, Not Your Org Chart The best discovery question isn't about strategy or structure. It's 'walk me through what you actually did yesterday', because the drudgery is the spec. read the post → Earlier · dev The Return of Custom Software For twenty years the answer to 'should we build or buy?' was almost always buy. AI just quietly flipped that back, and most companies haven't noticed yet. read the post → Earlier · work Guard Your Cash Like the Little Guy You Are You are the last vendor in the chain and the first to get squeezed. Price like it, bill like it, and build your way out of trading hours for dollars. read the post → Earlier · work Once You See Too Much, You Go Independent Nobody leaves a stable job because a spreadsheet told them to. They leave because they saw how the thing really works and couldn't un-see it. read the post → Earlier · work The Agency's Original Sin Sales promises the moon. Delivery is handed a telescope and a deadline. This gap isn't one bad company. It's built into how the whole industry pays itself. read the post → Earlier · everything-else I Didn't Mean to Start a Blog I don't think my opinions are better than anyone's. I just wanted to catch them before they evaporated, and writing is the only way I know to find out what I actually think. read the post →
01 / 24
