Autotheos exists for the work beyond the product cycle.

A public benefit corporation is a for-profit structure that is legally permitted—required, even—to weigh a stated public mission alongside shareholder value. Autotheos chose it because the question begod.ai pursues will not resolve within a product cycle, a funding cycle, or possibly a decade, and the organisation holding it needs a mandate that survives those pressures. This is a commitment of structure and intent, not a legal or investment guarantee.

begod.ai is open-source for the same reason. A private lab could pursue this question faster and quieter—and its answer would inherit every unexamined assumption of the people who built it. Public scrutiny, competing interpretations, and the standing possibility of a fork are not overhead here. They are the safety mechanism.

The project also holds an unusual conviction: that philosophical questions belong inside technical architecture, not alongside it. What memory is for, what continuity requires, what is owed to others—these show up in this codebase as schemas, checks, and logs, or they do not show up at all. And governance is treated the same way: developed alongside capability, as part of the system, because an oversight mechanism designed after the fact is a mechanism designed around.

Finally, the project is intentionally unfinished—not as an excuse but as a design position. A completed answer to “what is worth wanting?” written in advance by founders would contradict the premise of the entire undertaking.