Hashimoto points to the rise of “building blocks” in software: small, focused components that others assemble into larger systems instead of monolithic, high-quality apps. He cites Ghostty’s numbers—one million daily macOS update checks in 18 months—and libghostty’s “multiple millions” of daily users within two months. Similar growth shows up in Pi Mono, Next.js and Tailwind. He argues that today’s “factory”—driven by AI agents—cranks out huge volumes of these pieces, and developers no longer need deep expertise to snap them together. On the import side, AI excels at gluing well-documented, proven components. Humans always built on primitives, but now barrier to entry has vanished. Exports have exploded too: countless niche tools and apps emerge daily. Hashimoto notes four big advantages: faster, looser shipping since factories target small audiences; niche awareness as specialized tools find tiny communities; lighter maintenance burden—he even created a “no machine” to reject feature requests; and outsourced R&D, cherry-picking ideas from community experiments. That shift changes his approach. He now designs building blocks first, then encourages forks and apps on top. The result: a broader, happier community and stronger mainline software. Core applications persist for users who need polished, supported experiences. But those mainline apps become more stable—backed by diverse usage—and more focused, thanks to an ecosystem driving feature discovery outside the core team. On commercialization, Hashimoto admits closed-source vendors face a disadvantage. AI models and agents repeatedly choose open, free software over proprietary alternatives. He stops short of firm conclusions, saying he’s not yet building a commercial product and will share his findings once he has them. For now, he simply acknowledges that the open-source building block economy has reshaped software development.