Self++: Co-Determined Agency for Human--AI Symbiosis in Extended Reality

Why Read This from an HCI Perspective

This article shows that combining XR and AI is not just about boosting productivity—it also reshapes the user’s cognitive, behavioral, and relational context. In particular, the criteria of Transparency, Adaptivity, and Negotiability can be directly mapped to HCI design principles for explainability, control, and trust calibration. Its significance lies in letting us see both the risks of excessive automation and the potential for manipulation at the same time.

CIT's Commentary

The core of Self++ is that it treats AI not as a mere ‘tool,’ but as a coupled system that operates together with humans. In XR, this coupling touches both visual veridicality and behavior prompting, so the assumption that ‘kindness equals neutrality’ can easily collapse. Here, T.A.N. is not read as abstract ethical rhetoric, but as an interaction contract: users must know what has changed, support should scale down as growth occurs, and users should be able to roll it back at any time. That said, because role-based design is so powerful, we need to handle operational guidelines—such as when to stop intervening—and cultural differences more precisely to improve real-world applicability.

Questions to Consider While Reading

• Q.In XR environments where role switching is frequent, what notification approach preserves Transparency without breaking the flow?
• Q.When expanding Negotiability from personal settings to the level of collective agreement, what authority structure is actually the most stable?
• Q.How can we verify over the long term whether AI that supports Relatedness functions to strengthen relationships rather than replace them?