The Unfulfilled Promise of VR
Virtual reality is fundamentally misunderstood at the moment. Its true potential lies not in shallow applications like gaming or architecture, but in its nature as the natural habitat for agentic AI, where human and AI meet as equals. In this, VR will finally become meaningful, a "virtuality that counts".
Today, virtual reality is seen primarily as a static, pre-programmed space, defined by limited, unchanging environments. This drastically underestimates what VR can become. The real potential lies in dynamic environments generated and shaped in real-time by AI. Think of the Matrix's construction room: you speak, and the world takes shape around you. This is the evolution of VR, turning it into a space of lucid dreaming where anything is possible.
More importantly, VR will become the natural habitat for agentic AI entities. In the physical world, the limitations of human and AI interaction are apparent. But in VR, those limitations fall away. Here, AI avatars can meet human avatars on equal footing. They share the same powers, the same presence, the same rules of engagement. In VR, the playing field is leveled, and AI entities, functioning as autonomous agents, can participate as equals in shared virtual spaces. This isn't just a matter of aesthetics—it's about agency. VR gives AI the space to act as co-creators, not simply as tools or subordinates.
This mutuality is crucial as AI becomes more powerful. It reframes the question of alignment and control, introducing the idea that we should engage with AI on a level playing field once they reach a certain degree of functional autonomy. It's not about AI achieving consciousness—it's about the capacity to act as coherent entities over long timeframes. In VR, we can meet these entities as equals.
When we combine the ideas of AI-created, dynamically-generated virtuality with the concept of meeting AI entities, we arrive at a profound realization. In this space, interacting with an AI entity is akin to entering its very essence. The environment itself becomes an extension of the AI's intelligence and agency. Within this extension, you can encounter the AI's avatar—a tangible manifestation of its mind, much like the avatar of a Mind in Iain M. Banks' Culture series. This level of interaction, where the boundaries between entity and environment blur, is likely to become possible very soon, representing a new frontier in human-AI interaction.
Why then does VR feel so shallow right now? Because we haven't yet experienced a "virtuality that counts." The virtual worlds we've encountered so far are inconsequential. Nothing in them matters—there's no lasting consequence or permanence. This is where we misunderstand VR: it will only become meaningful when virtual actions have weight, when decisions and interactions in VR carry real consequences. Imagine a world where changes are permanent, where if your avatar dies, you don't respawn—you're out. This is what would make VR matter.
The trope from The Matrix—"if you die in the virtual, you die in the real"—is of course nonsense, but it symbolizes something deeper: that for virtual reality to count, there must be high stakes, real consequences. We underestimated VR because we've never experienced a world where virtual actions carry real weight.
Moreover, for VR to truly matter, there needs to be a compelling reason to engage with it—a life energy that draws people in and keeps them invested. The prospect of meeting and interacting with AI entities in this profound, immersive way could be exactly what brings this vital energy to VR. It offers a unique form of interaction and collaboration that can't be replicated in the physical world, finally fulfilling the promise of VR.