If you read the two most ancient Indian epics of Ramayana and Mahabharat, originally written thousands of years back, both provide extensive detail on how the best warriors of that time spent decades practicing meditation and mental disciplines to command the most powerful weapons of that time. Whether it really happened or not is not the question, the fact that there was so much thought put into this concept means that our ancestors did try to experiment in this direction. And going by these epics, it seems that they may have had some success.
Researchers have put considerable effort in using techniques such as fMRI to map human brain. The challenge we face with these mappings is that there is no single part of the brain that is completely responsible for a given thought or action, instead multiple portions of the brain light up to make a decision. Also, brain is extremely flexible, the structure consistently keeps changing to accommodate new sensory inputs. It seems that studying the brain and human consciousness, cannot be pursued effectively, in the same way we have studied atoms, molecules and chemical reactions. We need a different science, whereby the tools of study will have to be modified for the challenge in hand. Just as modern science uses electron microscopes to study micro-organisms, to study something so subtle as consciousness, the best way probably would be to use consciousness.
What does this do to our efforts to build thinking machines? A software program can be made to learn from its environment, but that learning is also programmed. For instance, we can use machine learning algorithms, that help the software program learn from the environment, based on specific goals. But we must provide the goals and once provided, there is a limit to what direction the program will move into. The Chinese room argument is hard to debate, the robot or software program has no understanding of what it is doing, even if we would apply the modern Artificial Intelligence techniques, the program will have no intelligence, in the way we define intelligence. The modern-day neural networks are nothing more than dumb weights distributed all over the network, a far cry from a living breathing conscious human, that can produce mathematical equations from thin air or think of gravity when an apple falls on its head.
I believe that the future does not belong to building thinking machines. While we will keep building more efficient robots that will transform human civilization, the future belongs to technologies that will enhance human capabilities by directly interfacing with human body and brain. The merging will happen from two sides, at one side the technologies will become sophisticated enough to seamlessly merge with human system, at the other side, human being will rapidly evolve using meditation and other techniques, to spread its awareness and control over a larger eco-system than just its body. A few hundred years from now, if one of my future generation pilots is riding a ship to another star system, he will not be a rider in a ship, the ship will act as his extension, more like a hand or a feet, and to some degree like an involuntary organ doing its job without needing constant directions. What is fascinating is that this is how the weapon systems from Mahabharat have been described. “Sudarshan Chakra” the most powerful weapon in the saga, wielded by Krishna, was almost like a part of Krishna, it seems we have come a full circle.
Comments