David Koelle 🌳

It's done. After years of work, Logic Islands: Deduction Game is live on iOS and Android. In an age of vibe coding and endless promises that you can build something in a weekend so you never have to think again, we did the dumbest possible thing. We took years to make something for actual human beings, to help them think more, not less. Every puzzle is made by hand and by obsession. I have been designing them for close to 15 years, slowly refining down to the core few hundreds that truly work, testing them on real people of every background and trashing the thousands that didn't make the cut. Each puzzle has solution walk-through explanations that we wrote by hand, on paper and whiteboards, then sat with players to see what landed and what didn't. Sometimes the most elegant (mathematically) solution wasn't the most illuminating, so we adapted. Some single explanations for one puzzle took days to pitch just right. 10 years ago I had notebooks full with ideas and puzzles for this game. I always had a lot going on, and I wasn't an app developer, so it felt too daunting. But I knew deep in my bones that it had to exist, and that I was built to make it, so it was only a matter of time. After getting hooked on the puzzles herself, my wife @NadiaGasmi convinced me we should build it together and build the version that gets it to the most people possible. 18 months ago, when she quit her job as a neuroscientist at the Cleveland Clinic to go all in with me, I knew it was really happening. I have always believed that one of the clearest marks of a good game (a game being just "voluntary attempts to overcome unnecessary obstacles") is that it resonates with people of wildly different backgrounds and ages. In early testing, in a given day: lawyers and engineers would ask if they could get their friends an invite, since they were hooked and wanted to share it; my (precocious) 10-year old niece would steal her dad's phone every morning and make him late to work because she wouldn't stop playing; and a few of my math PhD colleagues asked me to write papers about some of the stranger puzzles I was testing. This range of reaction, from such different people, gave us huge boosts and extra courage when we needed it. It was brutal at times. I had to keep working my 9-5 as well to keep the lights on while we built this, so I learned to love the evenings and weekends hammering away at it. Since I work largely on AI and math by day, it was interesting and refreshing to work on (what I came to see as) the opposite of AI by night. Especially now, when nearly everyone is claiming they can build something, without thinking, that will quietly strip away our ability and our need to think, and when too many people on social media seem bent on contributing to demoralization campaigns even when it doesn't serve them, we believe that a game that truly challenges you and makes you think more while having fun is extremely valuable. We have also watched the backlash as some big-name apps ram ai down their users' throats. So we are glad that we made the call at the start that we wouldn't hand over to ai anything we knew would be better if we did it ourselves, slowly and without shortcuts, every idea tested on real people, even when it cost us a great deal more time, patience, and expertise than we thought we had. Fundamentally, we made this game because it is 2026 and it is time to make the things that you feel in your core that only you can make: the things that must exist because you love them, that you love so much you know you can serve the people who use them for decades, the things that will help people have more fun, think harder, and feel better about the incredible thing it is to be a human being. Now go play. Link just below.