An evening with Dr. S.S. Mantha

On Friday, May 20, we had a very insightful session by Dr SS Mantha, former Chairman of AICTE. He explained some of the subtle nuances of Artificial Intelligence and how it is shaping individuals, industries, education and society. Dr Mantha also raised some thought-provoking questions and shared his perspectives on how AI will evolve in the coming years.

About Dr SS Mantha

Dr. S S Mantha implemented several e-governance initiatives and facilitating regulations in a six-year stint with AICTE. Earlier, a Professor of Robotics, Control theory and AI at VJTI, he now occupies the Emeritus Chair. He has more than 280 publications in National and International Journals and Conferences besides authoring three books. Dr Mantha is the author of NVEQF, now renamed the National Skill Qualification Framework (NSQF). He was member, Karnataka Skill Development Authority and advisor to the Government of Andhra Pradesh.

Currently, Dr Mantha is the CEO of MAHAPREIT Start-up Knowledge Centre, a subsidiary of the Government of Maharashtra.  He is also an Adjunct Professor at the National Institute for Advanced Studies (NIAS), Bangalore, an Emeritus Prof. at VJTI and Chairman of Technical Committee, National Cyber Safety and Security Standards (NCSSS), and IT expert, for the Government of Maharashtra.

Introduction

According to a famous Sanskrit sloka, things which others control cause pain while those which we control bring happiness. Dr Mantha started off with some basic questions. What is Atmanirbharta? What is AI? How are they connected? When things are so much dependent on each other, is Atmanirbharta or self-dependence at all possible? For example, the Earth is dependent on the Sun. The Sun is dependent on the Cosmos and so on.

Sometimes these dependencies can only be felt and not observed. There are virtual particles in quantum theory which deals with properties of nature at the scale of atoms and sub atomic particles. These virtual particles do not behave like normal articles and are short lived. Their presence cannot be detected directly but can be felt only through the impact they make on other particles in terms of charge, mass, etc. This places limits on the extent that one can control, ie be Atmanirbhar. We need science beyond the traditional boundaries to understand these deep concepts. We need both intelligence and forbearance.

Can AI exceed human intelligence or exist independent of it?

Human intelligence is a combination of complex cognitive fields, motivation and self-awareness. This combination allows us to learn concepts and apply logic and reasoning to find patterns and make decisions. We use a combination of different senses like sight, hearing, smell, taste, touch, etc. All these senses (along with vestibular senses, which help us to balance ourselves) together make the human world small. In the real world today, there are sensors which can measure all these parameters and AI can leverage this data.

The key questions in front of us are:
Will AI surpass human thinking?
 Can AI predict the future?
Will AI grow uncontrollably wiping out humanity?
Will there be AI singularity, ie AI takes over and controls human beings?

Predicting the future

We progress by anticipating and predicting the future. We understand the current conditions and then anticipate how the environment will evolve. Atmanirbharta (self reliance) involves solving problems by predicting the future.

Is it possible to predict the future? Quantum theory thinks otherwise. According to this theory, we cannot predict the future as the world is random and chaotic and not in a steady state condition. But AI can predict the future by studying huge volumes of past data, analysing it and then finding patterns. Such predictions are time consuming and not easy for human beings who do not have a similar level of computational abilities.

AI can mimic human cognitive functioning. But it can also do an unbiased analysis which will be difficult for human beings.AI can do the heavy thinking on our behalf. Machine Learning (ML) is a great enabler. It is a subset of AI. Actions within AI happen due to ML. In ML, the machine learns on its own, when fed with more data. There is no explicit programming.

AI will influence the way we think, work and act. It will have an impact across sectors. AI can make predictions using much data. This way, it can facilitate Atmanirbharta.

How AI is controlling our lives

Our daily routines are influenced and indeed controlled by AI. When we wake up, we look at our smart phone to unlock it. The biometrics based recognition system is based on AI. When we visit a social media website, AI algorithms decide what we should see and with whom we should become friends. The mail we use has AI to type out the message and also filter out spam. Antivirus software also uses AI.  Google search is driven by AI and so are the ads which pop up. On the way to the office, AI provides us directions, traffic patterns and weather information. Cars assisted by AI for driving are not too far away. Smart thermostats maintain the right ambience for us. When we use a debit card or credit card, AI checks whether there is fraud. Amazon not only has a great AI based recommendation engine but it also has developed an anticipatory shipping algorithm which can ship goods to us even before we order them. In the evening when we come home and want to relax, the Netflix algorithm recommends what movie we must see. Its recommendation will even take into account the time of the day.

Machine Intelligence vs Human Intelligence

Can AI outsmart human beings? Machine intelligence may catch up with Human intelligence (This is what we call Human Level Machine Intelligence) with a probability of 50% in the next 45 years and with a probability of 10% in the next 9 years. Elon Musk has mentioned that Machine intelligence will overtake Human intelligence in the next five years.

AI has become all pervasive. It manages our daily activities and influences outcomes. The locus of control is the determinant of happiness. AI takes control away from us.

Instead of becoming self-dependent, we are becoming dependent on AI. So the question we need to ask is: Are we becoming Atmanirbhar (self dependent) or AI Nirbhar (dependent on AI)? In the mad race to mine data and predict the future, we must sit and pause and understand our roots, i.e. our consciousness. The human mind or the human consciousness or the human intelligence is unconditional, unchanging, unfathomable, unique and imperturbable. We must hold on to our consciousness. That cannot be replicated by AI. Machine intelligence cannot help us cross the metaphysical bridge. In short, the highest level of human intelligence can never be surpassed by AI.

Q&A

On how AI is impacting different sectors

Agriculture: It is important to predict the uncertainties and provide insights to farmers, for example when to sow and harvest. Algorithms can predict when the monsoons will begin, when there will be sunshine and provide advice on when to water the fields based on the soil moisture.  The entire supply chain often described as farm to fork can be made more intelligent using AI.

Defence: Warfare today is driven by AI.

Healthcare: AI can be used to predict disease outbreaks like the pandemic. Doctors can be trained to use AI to improve surgical procedures and to make more accurate diagnosis. This is already happening in Amrita Hospital, Coimbatore.

Banking: AI takes care of KYC and various other activities and processes.

Manufacturing: AI is driving Industry 4.0, making machines intelligent and talk to one another.

Chess: AI can help chess players improve their game. Using AI, they can think two or three moves ahead on what they should be doing next.

On the use of AR/VR/Metaverse

AR/VR Applications are becoming common across sectors: education, medical robotics, entertainment. But the technology is still expensive and not affordable. With the attention span of students, no more than 17-18 minutes, classes should be gamified using AR/VR.

In the ICFAI Online MBA program, various kinds of animated content are being used. Gamification is being used to improve student engagement.  Technology is being used to personalize the learning and map it to the student’s situation. More uses of AR/VR are being explored.

On the employability challenge

Automation means that many entry level, low skilled jobs are going away. Earlier, we had a pyramid structure. We had many low skill jobs at the bottom and a few high skill jobs at the top.  But now even at junior levels, there are fewer jobs and higher order skills are the need of the hour. So the pyramid is being replaced by a pipeline. So reskilling/upskilling is the need of the hour.

We cannot fight change. We can only equip ourselves to embrace change. Many jobs today did not exist earlier. The jobs of tomorrow do not exist today. So we need learnability and cognitive flexibility so that we can remain relevant in different environments. For example, the Gig economy may demand a different set of skills.

On the digital divide

The divide between the rich and the poor is not only due to technology. The issues are more complex and have to be thought through. Every technology provides opportunities. But the way it is used can also pose challenges.

On the slow pace of adoption of AI by industries

Companies are typically bogged down by their ongoing operations, their production schedule, sales targets, etc. But they must remember that the power of technology is growing exponentially. Those companies which do not embrace AI will find themselves left behind.

On using AI for upskilling and reskilling

AI can improve teaching in various ways. It can identify the unique needs of learners and personalize the training. AI can facilitate the development of new content that is intuitive and interactive.

On the changing role of faculty

Faculty should move beyond monologues and be more of mentors and guides. They must create a context around the concept. Teaching is a performing art. Teachers should know the domain, have good presentation skills and be able to establish an emotional connection with the students. Teachers should be very actively involved in content creation and curation. In fact, if faculty pair up, it will be even better. One teacher can develop the content and the other can review, interpret, ask questions, and visualize how it can be used in the classroom. In the online mode, the effort required from the faculty is much more.

On building the technical capabilities of the students

It takes time to develop competencies. The four years of engineering curriculum only provide a basic understanding. Students should supplement it by being open to other learning opportunities available outside the classroom including internships. Students should try to understand how industry works and how decisions are taken. They must look around for learning opportunities and keep learning.

On using AI to improve employee retention

Employee retention is a function of several factors: aptitude, working conditions, compensation, career progression, learning opportunities, etc., Algorithms can personalize the working environment for employees. AI agents can be used to handhold employees and understand their concerns and aspirations.

On the trustworthiness of AI algorithms

All algorithms are thoroughly tested before being deployed. So we should not start with the assumption that they will fail. But algorithms are based on a certain data set and can misbehave when encountering new situations. Then we should make the required tweaks.

Note: As Prof Kartik Hosanagar mentions in his book, “A human’s guide to machine intelligence”, as long as we have some control, even minimal, trust in an algorithm is significantly enhanced. Therefore, it is a good idea to introduce an algorithm in an advisory rather than decision making role. It is also a good idea to introduce simple features that give people a feeling that they are in control. When automated elevators were introduced, people were scared to use them in the absence of an operator. What reassured lift users was the introduction of a red button that said “Stop”. The button was no big deal. In fact, if someone pressed the button, they were directed to use the phone and speak to a remote operator. Yet, the Stop button made users feel as if they were in control. A thought provoking session by a top academic leader of the country.

We thank Dr. Vedpuriswar for bringing out the highlights in the form of this note