Following on from my last post: if AI can write a polished "5 tips" post in seconds, the only content that still sets you apart is content that shows your judgement. Here are four formats that do that, built on a ladder I use called the Four Rungs of Expert AI Work. Each one shows a level of judgement a generalist, or an AI, can't fake. 1. THE GRADE Take something everyone in your field accepts, an AI answer, a popular template or standard advice, and mark it like a reviewer would. Hook: "I asked AI to [common task in your field]. It got 3 things wrong. Here's what a [your role] would spot." Why it works: it shows you can tell good from good-looking. 2. THE REFERENCE ANSWER Don't share the template. Share the version you'd actually use, and explain why each line is there. Hook: "Everyone shares the template. Here's the version I'd send to a client, line by line." Why it works: it shows what excellent looks like in your world, not what's average on the internet. 3. THE CHECKLIST Share the checks you run before you sign off on your own work. Hook: "Before any [deliverable] leaves my desk, it has to pass these 5 checks." Why it works: it turns your standards into something people can save and reuse. Checklists get saved, and saves tell the algorithm the post is worth keeping. 4. THE EXCEPTION Find where the standard advice breaks. Hook: "This advice is right most of the time. Here's exactly when it will cost you." Why it works: only someone with real experience knows where the edge cases are. That's the clearest signal of expertise you can post. Example from my field (finance): Grade: "I asked AI to review a cash flow forecast. It missed the one assumption a CFO would question first." Exception: "'Automate your invoice processing' is good advice, until a supplier emails asking to change their bank details." Your turn: drop your field in the comments and I'll reply with a hook for format 4, the exception post, which is the hardest to come up with on your own.

Hi everyone, The LinkedIn posts that win clients in 2026 show judgement, not activity. AI can now write a polished "5 tips" post in seconds, so polish no longer sets you apart. What buyers can't get from a machine is your standard: how you decide what is good, what is risky and what is wrong. Two pieces of research point the same way: Edelman and LinkedIn's 2025 B2B Thought Leadership Impact Report surveyed nearly 2,000 professionals. 86 per cent of "hidden" decision-makers (the people who shape a purchase without ever talking to sales) said they look for bold ideas that challenge assumptions. Stanford Digital Economy Lab, August 2026. Employment of 22 to 25 year olds in the most AI-exposed US occupations stands 19 per cent below where it would have been had it kept pace with less-exposed peers. Experienced workers in the same jobs show no comparable gap. The market is protecting experience and the judgement that comes with it. So put your judgement on the page. Here are three formats, adapted from the "four rungs" in my new book, each one showing more judgement than the last: 1. THE EXAMPLE POST Show what excellent looks like in your field. Put a weak version and a strong version side by side, and explain the difference in three lines. 2. THE RUBRIC POST Publish your standard. "How I judge whether a [proposal / hire / forecast / landing page] is ready: 5 checks." This is the format that gets saved and shared, because readers can use it straight away. 3. THE CAUGHT-IT POST Tell the story of a confident answer that was wrong, whether it came from a tool, a report or an "expert", and how you spotted it. This is the hardest of the three to fake. It is also the post that makes a buyer think: I want that person checking my work. Try one this week and drop the link below. I'll comment on as many as I can. If you'd like the thinking behind this, my new book The AI-Powered Future of Work (Book Two of my Human-Led Leadership Series) is free on Kindle until 5 October 2026: