OpenClaw just shipped a massive 2.0 update built by 933 contributors (569 first-timers) with over 16,000 pull requests—roughly half of all pull requests ever merged into the project. The seven-week development cycle was unusually long for a team that normally ships every day or two, but necessary because the codebase and shipping process both outgrew themselves as the team expanded. They wanted to avoid the classic trap of pushing out a broken update that would wreck existing installations, so they spent the extra time making sure the release worked cleanly for both fresh installs and upgrades. The update overhauls installation and the browser experience first, since those are where people actually encounter OpenClaw. New users now start with what they already have on their computer—existing ChatGPT or Claude subscriptions, API keys, local models—cutting out unnecessary setup friction so they reach their first conversation faster. The browser app got rebuilt as a real first-class citizen rather than an afterthought, letting people keep configuring things as they go, return to ongoing work, and watch live activity when they want to. The philosophy behind this release treats OpenClaw as something that starts simple and grows with you. A basic workflow might just monitor your kids' school emails and send Telegram alerts. From there, tasks can expand—your Claw could fetch your iPad receipt when your brother asks about it—without becoming harder to use. The team discovered this pattern while building the release itself: they needed shared cloud sessions so multiple people could collaborate on the same Claw task without losing context, turning it into multiplayer software. Since OpenClaw is open source and not locked into any single AI provider or company, users actually own what they build rather than just renting access to someone else's design.