Musings on Automation: Can Robots Think?
Published on : Wednesday 05-10-2022
There is a way to teach a machine to perform reasoning without supplying the procedure itself, says PV Sivaram.
Anand was in a high state of excitement. There was a spring in his step as we started on our regular Sunday morning walk. I kept quiet, waiting for him to open up the conversation. Anand couldn’t hold himself longer, and burst out – Robots are so intelligent! I asked him, what makes you think so? The gist of his excited outpouring was – He had visited an exhibition, the Automation Expo 2022. (This is India’s flagship annual show on Automation, but thanks to Covid, held physically after a gap of two years). There were hundreds and thousands of stalls and exhibits, which were full of equipment and information, he gasped. But the highlight for Anand, the exhibit which impressed him most, was a robot named Shalu. How did Shalu manage to impress teenager Anand? Not by any physical assets, as you might immediately jump to conclude. No, Anand was impressed by the intellect of Shalu who is not so attractive physically. Shalu can speak many languages and reply to questions, even on engineering subjects. This convinced Anand that Shalu was indeed intelligent. This is a common misconception, and I took it upon myself to clear this.
The topic is highly relevant today and in future because of the increasing deployment of Artificial Intelligence, Machine Learning and similar applications. These technologies are bound to have a profound impact on our work, home and economy. It is worth spending some time on this subject.
What makes a person or a robot intelligent? Commonly, we hold that intelligence is an ability to solve many different problems, answer questions on a variety of topics. This is how students are evaluated at various exams leading to certification, graduation and perhaps employment. By this measure machines such as robots are intelligent. Machines have different ways of learning – or rather, we have different ways of teaching them.
One method is, the machine is given a series of problems, and taught the correct answers to those problems. When the next problem is thrown up, the robot searches through its memory to find a matching question, and tries to apply the same answer or procedure to the new problem. It can also mean applying the techniques of the earlier solution to the new problem. We classify this technique under Artificial Intelligence (AI). This sort of AI is a very useful hand-maiden for most fields. As the accumulated knowledge of problems and solutions grows, a human feels challenged to remember and find a closest match out of this database. This is the task most suited for a machine to outshine a human.
This means that the machine can never become the master, right? Anand interrupted. Not so, I countered. There is a way to teach a machine to perform reasoning without supplying the procedure itself. This method is called neural network method or machine learning (ML). Here we supply a large data set of problems and solutions, and the machine works out the procedure to go from problem to solution by itself. Just as different persons have their own way of arriving at a solution, the machine too charts its way. This way may be quite different from the way any human would go about solving the problem. Most of the time, humans can’t even decipher how the machine solves the problem. This looks, from the outside, that the machine is mimicking the way a human brain works. Hence the technology is replete with terms out of human biology like neural networks. But, Anand, be sure the way machines work is still mechanistic. Even in ML, the machine learns only out of the dataset supplied to it and cannot go beyond it. Human brain has its own set of tricks and magic.
Intelligence is an ability to deal with situations and problems which had not previously been encountered by the individual. Human beings call into play traits and capabilities which humans themselves do not fully understand – like humour, intuition, guesswork and so on. Hence, my dear Anand, I would still hesitate to call a machine intelligent.
Anand had many questions after this introduction. I thought it was good for him to work out his answers before being totally spoon fed. Hence this discussion is adjourned to a next session.
PV Sivaram, Evangelist for Digital Transformation and Industrial Automation, is mentor and member of steering committee at C4i4. He retired as the Non-Executive Chairman of B&R Industrial Automation and earlier the Managing Director. He is a past President of the Automation Industries Association (AIA). After his graduation in Electronics Engineering from IIT-Madras in 1976, Sivaram began his career at BARC. He shifted to Siemens Ltd and has considerable experience in Distributed Systems, SCADA, DCS, and microcontroller applications.
Sivaram believes strongly that digitalisation and adoption of the technology and practices of Industry4.0 is essential for MSME of India. He works to bring these concepts clearer to the people for whom it is important. He believes SAMARTH UDYOG is nearer to the needs of India, and we must strike our own path to Digital Transformation. Foremost task ahead is to prepare people for living in a digital world. He is convinced that the new technologies need to be explored and driven into shop floor applications by young people. We need a set of people to work as Digital Champions in every organisation.
