Much a do about AI-the truth and the half truth

Photo by Steve Johnson on Unsplash

We are in a Goebbels’ world. Lest we forget, Joseph Goebbels was one of the most trusted sides of Hitler who masterminded propaganda campaigns using radio, films and press to promote the supremacy of Nazi ideology and brainwashed many Germans into believing that they were the superior race and Nazi was the best thing to happen for Germany. Something similar is happening in the present times. Take the case of Artificial Intelligence(AI). Goebbels’ canon that an oft repeated lie becomes a truth is being used to make us believe that AI is superior to the human brain and is soon going to replace Human Intelligence(HI) in all endeavors of life. And there are many who are willing to believe this half truth. The truth, however, is different. There is a need to understand that behind all this hype and hoopla about AI is a systemic campaign driven by the handful of rich and the powerful billionaires who are eyeing and vying to become trillionaires. Who knows some of them might already have become one.

Coming back to the issue, the fact is that AI is likely to remain limited to a narrow domain where it certainly has a far greater edge over human intelligence. But for that matter even a $20 pocket calculator is much better than humans when it comes to mental arithmetic. Talking about Deep learning, it is a pattern recognition technique and patterns are everywhere in nature. But even in this case it is to be noted that this ability is limited to select mental tasks. Most jobs involve many tasks even if it is not obvious. AI is making rich richer and poor poorer. Not only that, it is also making humans dumb as they are gradually losing their ability to think and use their creative faculties. But no matter how loud and vociferous the champions and the lobbyists of AI may argue, the challenges still remain. The same that were there initially. It is not about making a high powered computing machine. That has been done and done quite effectively. But the critical issue remains. That of designing the thinking machines. Thinking in all aspects of the term.

Let us understand this with the analogy of a human child. It develops slowly, gains experience through continuous interaction with the environment that is dynamic and ever changing at every infinitesimally small unit of time. However, the AI of the computing machine has a packaged and fully developed form of this intelligence. It does and may acquire new abilities but they are no match for the human thinking which is always interacting with the environment through its sense organs. This development is a continuous ongoing process that comes naturally. The skills and knowledge that humans can acquire quite effortlessly and sometimes unconsciously, are too difficult for the computers to do. Even a three year old child moves around his or her house without bumping against the furniture or the walls and doors. Though not apparent, the child is performing a very complex task- assessing its distance from various objects, working out on the path that it wants to follow and getting full knowledge of the space that is to be negotiated. It calls for controlling the movement, the speed and the direction so as to adjust to the surroundings. Moreover, it has the ability to instantly rework on all these parameters in case somebody or something unexpectedly comes in the way. The filtering mechanism of the brain enables the child to ignore many of the details and factors that may be a part of the environment but not relevant for the activity to be performed. To some extent, more advanced robots and the driverless cars have been infused with this ability but the human brain is still superior on this count.

Human thinking is based on experience based knowledge of the real world and the inbuilt ability to use it in a meaningful manner. Another interesting feature of human experiential learning can be summed up in this quote of Aldous Huxley who says that experience is not what happens to you, but rather what you do with what happens to you. So the same experience may lead to different outcomes in different people. The golden rule in psychology is that human beings are similar but not the same. It is this truth that gives human beings an edge over AI. AI faces the problem of designing a program which represents all this experiential knowledge in the memory chip to enable it to recall the relevant possibilities to be used as and when required. Attempts to design such systems were made some five decades back, but those expert systems as they were called, had their limitations when equated with human thinking, which could work on a much wider scope. Expert systems, thus, worked well when they were designed to encompass life’s experience in neatly defined narrow areas of specialization like medicine.

It has to be understood that AI driven systems could replicate human thinking, though not fully, in such specialized areas where the body of knowledge and experience is highly structured to be manipulated through a standardized set of instructions. AI can help a doctor better his diagnostic ability but that may not necessarily lead to more effective clinical sense. In the areas which are more fuzzy and flexible, writing the algorithm that may work, is rather difficult. All possibilities of life and their outcomes cannot be written in a program. AI can account for the logical and the rational but it will be clueless about the emotional that is erroneously understood as irrational. AI can have a mind, but it cannot have the heart and the gut. We all know that gut feelings that bring about intuitive decisions are often inexplicable. But they work nevertheless. True, a robocalypse is lurking round the corner but humanity has faced many an apocalypse and survived. This too will pass. Our consciousness and our conscience are two of our greatest strengths that no technology can acquire. Psychology will always score over technology.