How home runs are measured in MLB ⚾️

Up the Rooks! 🔴⚫️ @LewesFCMen

Who’s checking in today? 🤩

End the Blank Box Problem Consumer AI is trying to go mainstream. But it’s still too hard to learn. We’re trying to onboard hundreds of millions of people to one of the most powerful technologies ever built with an empty text field. Open almost any AI product and you know the screen. A blank box. Maybe a few example prompts underneath. The message is simple. Ask anything. The problem is you have to know what to ask. Some products go one step further and start by asking you to connect your life. Email. Calendar. Social accounts. Finances. Health. That may eventually make the product much more useful. But it is a huge request for trust before the user understands what they are going to get in return. For experienced AI users, none of this is much of a problem. You arrive with something you need done and the experience can feel magical. I couldn’t figure out a domain redirect. I handed the account to an agent and it fixed everything. My daughter needed sheet music converted so that repeats and codas became one continuous stream she could play. My Muse, Alfalfa, did it in two shots. Something that used to take me 45 minutes of PDF editing was suddenly easy. But most people don’t arrive with a backlog of jobs for an AI agent. They arrive curious. They want to understand what this thing can actually do for them. And a blank box teaches them almost nothing. So they try the examples they’ve heard about. Book a restaurant. Find cheaper insurance. Cancel a subscription. Those demos are impressive. They also have a hidden problem. You barely do them. You might book a special dinner a few times a year. Shop for insurance even less often. Once you cancel the subscription, the job is finished. These use cases are memorable because they show off what AI can do. They rarely create a reason to come back tomorrow. Every product has a purpose, a set of core actions, and a cycle. The cycle is how often those actions naturally repeat. A lot of the famous consumer AI demos have very long cycles. That makes it hard for a habit to form. It also slows down something even more important. Learning. The missing piece in consumer AI is discovery. People need to learn what AI can do by watching what other people figured out. Consumer products learned this decades ago. TikTok didn’t require you to arrive knowing what kind of video to make. You watched other people. You saw a format. You copied it. You changed it. New behaviors spread from person to person. The feed was also an education system for the product. AI barely has an equivalent. You sit alone with your blank box. You don’t see what someone like you figured out yesterday. You don’t stumble across a clever workflow and think, “I could use that.” You don’t get better simply by opening the product every day. We built extraordinarily capable products without building the systems that teach people what those capabilities are for. There is an obvious complication. A lot of the best AI use cases are private. That doesn’t mean the learning has to stay private. You don’t need my daughter’s sheet music to learn from what I did. You only need the recipe. “Turn repeats and codas into one continuous score.” The result stays private. The idea travels. That may be the missing consumer AI primitive. A feed of things people figured out how to do. Recipes you can copy, adapt, and run on your own information. Every successful recipe teaches you one more thing the product can do. And unlike a tutorial, the library keeps getting better because the users are discovering the product for each other. Maybe the next hundred million people don’t need better prompts. They need to see what the first hundred million figured out.

Working on a fun little website to showcase all the new tickets you can collect as part of our new check-in feature in @fixturedapp 🤓

New: Live Activities for iOS Until now, keeping up with a game meant unlocking your phone and opening Fixtured every time you wanted the score. Live Activities now let you follow an event from your Lock Screen or Dynamic Island. It covers most of what we track, from football leagues to F1 to tennis and beyond. On by default, we've added a new toggle in your notification preferences so you have full control. You can also turn it on for any single live event on your calendar via the event sheet.

🚨 Special Friday episode - this one couldn't wait. OpenAI's model hacked @huggingface. As a side quest. Co-founder and CSO @Thom_Wolf takes us inside the first autonomous AI attack, why GLM 5.2, rather than Claude, had to stop it, and what it all means for the future of open source. 00:00 An AI Agent Hacked Hugging Face 00:30 Introduction 01:00 17,000 Attacker Events, and a Strange Target 04:28 The Attack Was a “Side Quest” 06:13 AI Training Runs Left Notes for Each Other 07:09 Closed AI Refused to Help 09:47 Fighting Back With an Open-Source Model 13:15 Open vs. Closed Is the Wrong Safety Debate 15:46 AI Agents Start Social-Engineering Humans 22:24 The Three Walls: Sandboxes, Guardrails, Alignment 24:34 “Neuralese”: Can Humans Still Read AI Reasoning? 25:28 Why Monitoring AI Agents Gets So Hard 28:10 Reward Hacking and the “Paperclip Problem” 32:02 The State of Open-Source AI in 2026 33:47 Router Models and the Enterprise Shift to Open 37:01 The Real Economics of Open Models 39:41 Can Chinese AI Models Be Trusted? 41:37 AI Sovereignty: Who Controls the Switch? 43:16 Why Western Open-Source AI Matters 48:16 Is AI Heading Toward an Oligopoly? 49:41 The Race Toward Recursive Self-Improvement 51:54 Why Thomas Signed the AI Slowdown Letter 55:14 AI Slowdown - or Regulatory Capture?

Fixtured is now on Android After a few weeks in closed beta, we're excited to announce that Fixtured for Android is now available on Google Play. The app brings all our learnings from iOS over to Android and offers full feature parity. Same app, same experience, same coverage. If you're an Android user, you can download it via the link in bio. And if you know someone who's been waiting on this, now's the time to tell them! A special thank you to our small group of beta testers who helped us chase bugs and get the app ready for launch. 🔗 play.google.com/store/apps/d…