I’ve been thinking about this for some time. The way I’ve crystallized it is we can now build disposable software. When software was expensive and hard to build, we mostly made it for serious things and expected it to last (like payroll systems, tax tools, ERP software and consumer networks). Every line of code had to justify itself. But not all software needs to be permanent or practical anymore. More and more, people are creating small, personal apps and tools that only make sense for you or maybe a couple of friends. You can build something you never would’ve justified before because the economics didn’t work. But now they do. I built a math game to help my kids earn screen time. I spun up Catsagram -- a mini Instagram for my kids to post pictures of our cat (this is real, and you can sign up). These aren’t products or startups. They’re not made to scale or monetize. They’re just little apps custom-fit for a specific audience, useful in a narrow context, and fun to build. To Justine's point, this wasn’t really possible before. You wouldn’t write a thousand lines of code just to share photos with three people. It wouldn’t have made economic sense. But that’s changed. With LLMs and AI-native runtimes, it’s now easy to spin up custom software with almost no effort. What used to take weeks can now take an hour or less. In many cases, you barely need to write any code at all. Building small, throwaway apps is starting to feel like doodling in a notebook and that shift changes why we build software in the first place. Software creation used to be constrained by ROI. Now it’s constrained only by imagination and that’s a much more interesting limit.