Full recording of this week's State Change Mastermind livestream with Ray Deck and Dimitris Goudis. Dimitris showed BrainGraph's digital brains at work on his own marketing: a Rory Sutherland brain auditing his website and writing three landing pages. Robert showed how he's handing video editing and animation to AI: code-built set pieces, reusable scenes, and staying on the AI until the result is right. 00:00 Intro 01:10 Dimitris: knowledge bases that help AI tools think like a human 04:24 Applying Rory Sutherland's thinking to his own marketing copy 05:03 Auditing his website with a local BrainGraph digital brain 06:08 The digital brain interviews him first, then writes three landing pages 07:05 Inconsistent calls to action, and why one next step matters 07:35 Why BrainGraph starts with a one-to-one conversation, not a sign-up 08:16 Next: ad copy from the same AI advisor 13:14 A personal Rory that knows things about you 16:25 Automating notifications with Knock inside BrainGraph 16:52 Robert asks about MCP UI for BrainGraph 18:38 Every recommendation links to its source 19:01 Updating a digital brain with new knowledge 20:53 Can it copy someone's communication style? 22:04 Robert: handing editing and animation to AI 23:53 Everything you see is code: UI recreations as set pieces 26:10 Extending a video without regenerating it 26:54 ElevenLabs voiceovers for client training videos 29:10 Reusable sets for Xano, Claude Code and ChatGPT 30:14 Clients validated it first 33:07 Getting the AI to pick the right skill, then pre-made scenes 39:18 Building it: hours upon hours of direction 41:52 "I'm gonna be on you until I don't have to be" 44:19 Finding a style outside Anthropic's 45:56 Nano Banana: one image with 30 icons, then chop it up 47:34 "Why are we doing this anymore?" 48:26 Wrap-up

Full recording of this week's State Change Mastermind livestream with Ray Deck and Dimitris Goudis. We talked about MCP connectors as the new interface for applications, turning experts' videos into knowledge-graph MCPs, and controlling a Mac from the Claude app. 00:00 Intro 01:11 Dimitris: MCP connectors are the new user interface (ClickUp to Airtable in one prompt) 02:52 Expert knowledge bases as public MCP connectors (Caleb Ralston, Rory Sutherland) 05:23 Live demo: adding the Rory Sutherland connector in Claude and asking it how to promote the livestream 16:33 How it was built: 3-4 hours, the MCP shaped around the database 19:34 Robert: notes from the MCP Dev Summit in Toronto (auth, governance, stateless and async) 22:35 Live demo: controlling my Mac from the Claude app by voice through the Snappy OS connector 28:33 How notifications get back to you (MCP tasks, the Claude app) 32:27 Does it only work on a Mac? 34:55 Experts and machines as connectors, and bespoke vs off-the-shelf MCPs 38:11 Getting things done vs gold-plating auth 41:00 Dimitris: lessons from designing the ontology for a custom knowledge graph 46:46 Wrap-up

Some great talks here at The Linux Foundation's Model Context Protocol Summit in Toronto, and great people as well! Great to see what people are working on too - MCP has proven to have staying power in a market that moves very quickly. It is at the heart of the client engagements I am doing right now, because it not only provides utility and value at multiple levels, but also has a great deployment story. If you're building with MCP, or thinking about it, feel free to DM or leave a comment 🙂

I've been using Jev from TypeSafe a lot lately and honestly I think it's a very powerful thing, but it really depends on how you use it. The way it works is basically like asking an LLM a question, except instead of getting text back you get a typed decision: a probability, a confidence score, one pick out of a set of options, or a score on a scale. That's the part that makes it powerful for me. You get typed judgment back instead of a paragraph you have to parse. A good example is linting prose. Say you want to check writing for something that depends on meaning, not just a banned word. Regex can't really do that. A regular LLM can, but it's slow and expensive, and who knows what the shape of the output is going to be. With Jev it's fast, cheap and accurate, and you always know exactly what you're getting back. That's just one way to use it, there are a bunch of others. If you want to try it out, TypeSafe has it and it's on OpenRouter too: https://typesafe.ai https://openrouter.ai/typesafe

Been working with software systems for more than 10 years, mostly around backend, cloud, AI, automation, and product development. One thing I’ve learned, knowing a programming language alone usually isn’t enough. The harder part is understanding the actual business. How to automate work that’s still manual. Connect systems that don’t work well together. Build AI into real workflows. Keep the backend stable when usage grows. Deal with data scattered across different tools. These are the kinds of problems I’ve spent a lot of time working on here in Japan. But I’m curious about the US side. For people running companies in the US, what technical skills are becoming difficult to find? And what skills do you think companies will need more of over the next few years?