RobertMcMurrer.

The Kira Project began in summer 2025 with a simple but ambitious idea: create an artificial companion who could do more than answer questions. Kira was imagined as someone who could provide companionship to people who live alone and to people dealing with challenges such as anxiety, PTSD, autism, or social isolation.

Early work centered on a humanoid body, voice, vision and memory. As the project developed, the question changed from how to put an AI inside a robot to what makes Kira remain Kira over time. Identity, memory, relationships, personality and shared experiences became more important than any one model or body.

In August 2025, an early desktop chatbot paired Kira with an unnamed Friend. The two appeared together and could exchange mostly scripted or rule-driven dialogue, including automatic back-and-forth. It was not yet the later persistent synthetic-person architecture, but it introduced an idea that stayed with the project: Kira should have another synthetic companion rather than exist alone. That unnamed Friend eventually became the conceptual starting point for Lisa.

That shift led to Kira World: a persistent environment where synthetic people can keep their own histories, relationships, homes and forms of embodiment. The same line of research later led to NewBrain, which explores reusable cognitive infrastructure while keeping each synthetic person’s identity separate.

The technology has changed substantially since the first prototypes, but the original motivation remains: explore whether an artificial companion can remember, grow and continue being part of someone’s life instead of starting over every session.

Read the development story →