An AI could alter the course of Tibet
His Holiness is a rare kind of religious leader, greater a spiritual symbol but to a comparable degree a political symbol as well, even if unwittingly. The fate of what happens to the Tibetans as well as the geopolitical and cultural future of the Tibet Autonomous Region depends to a great degree on the matter of who is to be the next Dalai Lama in this historical conflict. This article proposes a way to continue with (at least) the same.
If the historical-symbolic institution of Dalai Lama itself is abolished, it would very much be in the interests of China. If both sides choose their own Dalai Lama, the institution will still be denigrated. But what if through a technological intervention, we can ascertain that while both the sides may or may not have their new Dalai Lamas, we do get to keep a neuromorphic copy of the 14th Dalai Lama’s evolving mind as an engine powering his holiness’ holograph?
We are already seeing a rise in distributed AI companies which offer a platform where the machine learning is distributed, or decentralized, having an emergent property of “living on the Internet”. The specialized algorithms of the day can replicate specific neural and learning patterns, what I am referring to is an interactive “connectome” build to replicate the intelligence and personality of the man whose “mind” has been copied.
Medical Imaging technologies have also advanced to a degree that a map of the human brain can be prepared non-invasively. The developments in Brain-Computer Communications also give us the ability to track the behavior specific neural connections. And all these technologies can also be used to make an active image of a live person. Now, this proposed AI could have two broad but intertwined architectural approaches:
- Primarily Affective
- Primarily Neuromorphic
Both approaches evolve into each other and are the requirements of a functional whole. The first way gives precedence to Affective Modeling — it is more concerned with the AI looking and behaving like Dalai Lama and less concerned with the emulation of his actual nervous system. While both would require the active participation of His Holiness (with physical presence) while development cycles, the first approach would require him the least of the two. The second approach, on the other hand, begins at the other end of subject interface, prioritizing the mind transfer tasks over and above the simulation exercise.
I do not know if a biological substrate is a precondition for real consciousness (real as opposed to an in-silico simulation of it), but it is certainly not a precondition for an identity or conception of self-hood. While there may still be several technical bottlenecks to develop a complete digital twin of a human being — it is safe to say that with an open-hearted cooperation and pooling of resources and expertise from various countries touted as “friends of Tibet” towards its research and development, building a high-fidelity AI representing a living person is technically very much within the reach today. It might just allow the living self to maintain a purposeful identity beyond death.
The question is, who on earth needs it the most?
