Subhash Kak: ‘It is wrong to assume that consciousness is just a computation’
Updated: 01 Mar 2019, 01:33 PM ISTNipa Charagi- The curriculum needs to move away from rote learning to critical thinking to deal with the AI challenge
- Each generation needs to reinvent itself and that cannot be done by mimicking other societies
Subhash Kak, who was recently awarded the Padma Shri for science & engineering-technology, says India needs to move from rote learning to critical thinking. Each generation needs to reinvent itself, says the Regents professor emeritus, Oklahoma State University, US, whose research covers the fields of neural networks, cryptography and quantum computing.
Kak, who has written 20 books, including six books of poems, says what sets humans apart from intelligent machines is awareness, and it is “wrong to assume that consciousness is just a computation".
“My research has led me through various pathways, some of which touch on ancient wisdoms and others on modern science," writes Kak, who is also a Vedic scholar, in his book The Circle Of Memory: An Autobiography (2016).
In an email interview, Kak, who is on the Prime Minister’s science, technology and innovation advisory council, says: “As a scientist, I have worked on the problems of foundations and I took the same inquisitiveness to explore the Vedic world and I discovered that at its heart it is deeply scientific. It speaks of the objective sciences, like physics, chemistry, medicine and so on, and a separate science of consciousness." He says that while ancient Indians did not clone babies, they did have the most advanced logic and mathematics of their time. Edited excerpts:
Are we entering an inflection point, where machines will be able to take decisions and self-learn? In that context, what will the future look like?
Machines are already taking decisions in pattern-recognition applications, such as machine trading and auto-pilot flying an aeroplane. They can also self-learn once the parameters of the application have been defined. Indeed, AlphaGo, the computer that defeated the world champion at Go, a strategy game more complex than chess, honed its game by playing against itself.
What is changing now is the breadth and scope of applications, like cars that self-drive on crowded roads. Since computers are more reliable than humans, it is inevitable that their use will only increase. There are also unprecedented dangers of thought control using AI technologies. For instance, AI-designed bots can spread false stories and do it in a manner that pushes out real news, and public opinion can be manipulated.
A lot of jobs will become redundant. What kind of churning would this involve in the education system, especially in India?
India faces a huge challenge regarding the quality of its education. India has some excellent schools and colleges but not enough of them, so there is a huge skills shortfall. The other aspect of this is how students are taught in classrooms. The curriculum needs to move away from rote learning to critical thinking. Educators have long been aware of it but not enough has been done to bring about qualitative change. Education is of paramount importance to deal with the AI challenge because the nation will have to create new kinds of jobs in place of others that have become redundant due to automation technology.
Countries like South Korea and Japan are said to be the most prepared for automation. How do you think India fares in this regard?
Yes, South Korea and Japan, as well as China, are far ahead of India in the field of automation. But given that the Indian economy has momentum at this time, the gap need not be too much of a handicap if right investments are made in AI technologies. India also has the advantage that much more needs to be done here to fix its infrastructure, so (there is) potentially more work for local companies.
Will future computers be able to mimic the human mind?
Computers will be able to emulate human cognition, which means that they will be able to match and surpass routine human cognitive operations. But in my view they will not be able to mimic the human mind. That is, they will not have awareness.
From an economic point of view, the absence of awareness does not change the picture related to jobs loss, but it makes AI machines a bit less of an existential threat. If machines were to become aware, they would surely do away with humans. Although I believe this will not happen, there are many scientists and engineers who are sure machines will become conscious. They are wrong to assume that consciousness is just a computation.
Should AI be regulated?
If you really think of it, AI is nothing but machine-based pattern recognition. Once one grasps this idea, it is clear that regulating AI does not make sense. However, there could be situations where the risks of using AI are high and society will have to grapple with the issue. For example, should a robot soldier be allowed to pull the trigger when there is a chance that civilians may be killed? Or should the computer on a drone be allowed to shoot or bomb based on stored algorithms?
Some regulation will be necessary to determine where AI could be used and who would be liable when things go wrong.
You emphasize the need for development of the science of consciousness. What do you mean by that?
Consciousness is awareness and all indications are that it cannot be explained by normal science. Still there can be deeper understanding of the phenomenon through fields such as physics, computer science, neuroscience. For example, it has been discovered that there is almost a half-a-second lag between the time of the decision in the neural circuitry of the brain and the conscious awareness of that in the individual’s mind. This is a mystery that needs further elucidation. There are also questions, such as if consciousness is not a material property, how does it interact with the brain, and how does freedom emerge in a world governed by scientific laws?
How does neuroscience align with Vedic wisdom?
Let me take just the question of the structure of the human mind. The Vedic view is that cognitive centres are independent, and it is in that sense that they hold up the inner sky of the mind. The independence of these centres has been established in neuroscience by results, such as that in certain brain injuries a person can lose the ability to read without losing the ability to write! This is called alexia without agraphia.
You say Indians are notoriously ignorant about the past, and that as much as you see into the future, you have to look into the past too....
There is the story of two brothers who were each given an old heavy shawl by their mother on her deathbed. One of them gave the shawl away to a beggar, for he didn’t like its feel and look, but the other, upon careful examination, discovered it had little diamonds sewn into it, which made him wealthy.
The past is like that shawl. It provides insights for skilful navigation through the unfolding present. But it needs work to find those diamonds that are hidden in its folds. Each generation needs to reinvent itself and that cannot be done by mimicking other societies. India will not know where to go by copying others; it needs to find its own way and for that it needs to know the past, not to obsess over it but to make peace with it.
There should be no confusing of poetic metaphors with facts. Ancient Indians did not clone babies or have aeroplanes, but they did indeed have plastic surgery and the most advanced physics, logic, and mathematics of their times. A course on the history of Indian science as a part of the school curriculum should address this problem.
You are a scientist, a Vedic scholar, a poet. How do these worlds come together?
In the West, C.P. Snow spoke of two cultures: one of the arts and poetry and the other of the sciences, and this division is seen in the way people live out their lives, choosing one or the other. I don’t believe in this division, for all creativity, whether scientific or artistic, comes from the same place in the heart.