Devlog #30: What we don't ship
Not every idea makes it. We kill features. We delay. We say no. This post is about the things we decided not to do and why. Transparency doesn't mean we ship everything.
We've killed or deferred features that would have been flashy but wrong for the world. Some were scope creep. Some were technical dead ends. Some were ethical lines—things that would have increased engagement at the cost of manipulation. Matchmaking, for example, can be tuned to maximize retention. It can also be tuned to exploit psychological vulnerabilities. We're building systems that could do both. We're choosing to do only one. Same with any future tools that touch creation or moderation: the same system that reduces churn can be used to manipulate behavior. We're not building that. We're building the version that serves the participant.
Control and creative governance
If we ever open the world to more user-generated content—generative tools, in-world authoring—we'll need guardrails. Not "whatever you prompt, you get," but design languages, safety policies, and performance budgets. A world can't accept arbitrarily generated content without risking exploits and collapse. We're not there yet, but we're thinking about it now. When we do ship creation tools, we'll ship governance with them. Provenance, rights, what can be built and how it's shared. Transparency doesn't mean we ship everything. It means we tell you what we're not shipping and why.