Today Artificial intelligence (AI) has got many applications in society. The form of AI known as “weak AI” consists of programs which carry out tasks in fields like robot control, medical diagnosis, e-commerce, remote sensing etc. Self-driven cars, translation machines, chatbots etc. also make use of AI. One question that may be raised is: Can AI in the future become actually conscious or self aware? Is it possible that after reaching a high level of sophistication, the AI system could attain self-awareness?
Although the AI may be able to simulate conscious behaviour or self awareness, it is highly unlikely that it can actually become conscious at any time. This means that an AI system can only pretend to be conscious, but it cannot actually be conscious in the way human beings (or even other living beings) are. Currently there are robots that interact with human beings and are very intelligent in their behaviour, and appear to be almost conscious. The current viewpoint of modern science is that the human mind involves nothing more than the biophysical functioning of the brain. The terms that are normally applied to human personality eg. consciousness, perception, meaning, purpose, and intelligence are defined in entirely mechanistic terms. Current science assumes that a living being is just a highly complex machine made up of molecular components.
But, there is huge difference in a highly advanced machine (AI) and a human being performing the same activities. For example, when a person reads a book, he becomes aware of various thoughts and ideas in the book, and is consciously aware of the plot of the story. They can also experience emotions like thrill, sadness, joy etc. All these are corresponding to the higher-order abstract properties of ink patterns on the paper. But actually, these abstract properties do not exist in the book itself, and the book itself is not conscious of its contents. When a robot equipped with a camera and image processing units “reads” the book, it only analyses the patterns of ink on paper and can carry out many tasks which are only mechanistic in nature: e.g. OCR output, count the words, translate the content, correct spelling and grammar etc.
This is a sign of intelligence, but the machine does not possess conscious behaviour as it cannot experience thrill, or joy or sadness in the way a human being can, because there is no actual understanding of the “plot” or storyline in the true sense. Hence to establish a correlation between the ink patterns and the conscious awareness of its contents, there must be a conscious entity with intelligence and senses who can actually read the book. But a machine which is an is made of entirely mechanistic construction cannot have this correlation, no matter what the level of sophistication. Then what is that conscious entity within the human being which makes that correlation possible? A mechanistic explanation of consciousness will lead to infinite regress. Hopefully, more research into a new realm of science, which is non-mechanistic in nature may provide us with answers in the future.
By: Prof. Mayur V. Gambhir
Assistant Professor, Electronics and Telecommunications