One fine day, inhabiting the portals of the library, a discussion begins, soon, passionate arguments are thrown around and very soon, it explodes into a conflict of ideas, and the resulting chaos threw me out of the library! (Well, yeah! The discussion was being the cause of anguish to the typically retentive, repetitive medical student who has his/her plans chalked out :P)
So, this discussion was about Artificial Intelligence, or to be more specific, why and how natural intelligence differs from artificial intelligence....?
Anyways, so, if you have reached this part of the blog-post, lets continue further....let us assume modelling a human brain, in all its complexity , is possible....then, let us try to answer some questions:
Q.1: What does the brain do?
Well, even today, after years of research, no body, can with certainty, clearly define and designate the brain’s functions.
The rationale behind designing an artificial heart/ leg/lung is that they have a explicitly clear and understandable function(s), that can be reduced to a simple mechanical bottomline for eg: the heart pumps blood, the leg is responsible for locomotion, and the lung facilitates gas exchange.
Conclusion:We are unable to, in gross terms, understand what consciousness is, let alone model it!!!
Q.2 :Are you the same person, today, that you were yesterday, and will you be the same person tomorrow that you are today?
Well, if you say, Yes-well, then, I’d say, No....Every Second is a learning experience for the brain- its internal circuitry being continuously changed by charging and updating!
But, even a computer can learn, right? But, again, would it be wrong to state that the scale of dynamism is very different? A computer, will learn and modify itself, only when, it has been “taught” to “learn” by predetermined rules (software)!
The brain, however, or rather, the biological neurons, don’t just change the strength of their connections, they grow new connections, and let old ones atrophy i.e. they change their physical shape => Hardware is indisguishable from software!
Roger Penrose and Deep ThoughtThe famous mathematician, Roger Penrose, in his inimitatable style, brought into focus, why the human consciousness cannot be modelled by a
conventional Turing machine.
Computers function by means of algorithms :
Step 1:
Step 2:
Step 3:
.............
............
However, the human brain does not work in this number crunching mechanical way!
Penrose pointed to the example of the chess computer, Deep Thought, which was confronted with a chess problem, in which it had only pawns left, apart from the king. The pawns were positioned in an impenetratable way, and all that a human player had to do to avoid defeat was to position the king behind the pawns. The mechanical way that Deep Thought was programmed, it took the rook, and thereby, lost the game.
Moral of the story: True, the game couldn’t be saved even by a seasoned chess player, but what Penrose tried to project through this example was that humans use more than just a fixed set of rules that are implicit in the problem to be solved.
Maybe, thats what Niels Bohr meant when when he admonished a student- “
You are not thinking, you are just being logical!”
The whole Brain, thus, can be thought of as a mechanism for reducing risk, by making more accurate predictions about the likely result of any particular course of action, on the basis of past experience and memories. And it is our consciousness that helps us select and edit what gets into our memories!
In sense, then, we can now, say, that we are our memories- accumulated, nurtured and pruned over a lifetime- this wonderful mechanism, offers us, some degree of freedom from neural determinism, and probably, and probably, makes us the unique person that we are!
Then, why did the Hypothesis that “we” are nothing but complicated machines arise? It is, to a large extent true, with respect to our bodies, because, regarding bodies as machines, has definitely enabled several breakthroughs in biology and medicine.
And, it wouldn’t be wrong to conclusively say, that it is only of limited validity, with respect to our minds.
Maybe, it was a similar hope, that considering minds as mechanisms can help us, one day, discover, what makes the unpromising, gooey, creamy mass floating in a liquid aka the brain “tick”?
And only when we are explicitly able to pin down how the brain makes chaos in order to make sense of the world, can we truly understand ourselves, our worlds, our worlds of sensation, thought, feelings and emotions....
P.S: If this post appears ambiguous, then, I would choose to explain by quoting Jung -
“I prefer ambiguous language, since, it does equal justice to the subjectivity of the archetypal idea and to the autonomy of the archetype!”
Afterthought:It definitely is a tough task to model the human brain, in all its splendour. However, if we, now introduce a generic term “neural nets” in this discussion, which, arguably is the closest we have got to modelling the brain in silicon, it is indeed possible to model the essential functions of the brain. As they say, it really isn’t necessary to build replicas of bird’s feathers in order to fly!