the JUP token launched a little over two years ago. since then, we've been through catstanbul, catlumpur, multiple acquisitions, and the team grew from fewer than 20 people to close to 200 people at one point. one of the biggest challenges for me wasn't technical. it was changing my role. when the team was small, decisions happened with a handful of people in a room. we could debate, build, ship, and iterate within days. as the company grew, my days became almost entirely coordination. getting alignment across teams could take weeks. every morning there would be another list of decisions waiting. there was no longer any uninterrupted maker's time left on my calendar. (pg's essay on maker's schedule vs. manager's schedule captures this feeling perfectly.) for a long time, i didn't realize what was happening. the first time you're deciding whether to spend millions of dollars, you think carefully about every angle. after making enough decisions like that, they stop feeling extraordinary. your brain starts optimizing for speed. pattern recognition replaces deliberate thinking. eventually, you're no longer thinking deeply about every problem. you're just unblocking the next urgent thing. looking back, i think i spent a long period operating almost entirely on autopilot. or maybe zombie mode is the better description. nothing was wrong externally. eventually, i realized what was missing. i had stopped learning. that was probably the biggest problem. i wanted something that would force me to become a beginner again. i wanted to think deeply again. i wanted to build again. i wanted to struggle with problems i didn't already know how to solve. that became GUM. building from first principles again has been one of the most enjoyable experiences i've had in years. i find myself reading constantly, learning something new almost every day, and having technical discussions that leave me with more questions than answers. the team makes it even more fun. i'm lucky to work with engineers who have built products that facilitate billions in trading volume, security researchers who constantly challenge assumptions, and former L1 engineers who think about infrastructure from completely different perspectives. every conversation feels like an opportunity to learn. for the first time in a while, i'm uncomfortable again. and i think that's exactly where i want to be. GUM's private beta launches next monday, july 6. i'm excited to start learning from users again.