Artificial intelligence, unarguably, has been transforming the lives of people in ways that were unimaginable till recently. While it has contributed immensely to improving efficiencies across systems, its unprecedented and unregulated pace has also caused anxieties in terms of deepfakes and potential job losses.
Dr. Raj Reddy, Moza Bint Nasser University Professor of Computer Science and Robotics at Carnegie Mellon University and Padma Bhushan awardee, has been a strong proponent of AI and the promises it holds for the world.
“AI is primarily to be seen as enhancing the mental capabilities of human beings. Anything you use your brain for, AI will be able to enhance what you do. And to me that is the main definition of AI, just like engineering as a field enhances the physical capabilities of human beings,” he notes.
Dr. Raj Reddy | Photo Credit: SPECIAL ARRANGEMENT
Dr. Reddy was recently in Bengaluru to deliver the inaugural lecture of Dr. V.S. Arunachalam Memorial Lecture organised by CSTEP. He spoke about the potential of AI to bridge literacy gaps and language divide and eliminate poverty and hunger with the help of AI-applied technologies that may aid in facilitating better food, energy and water security, medical care and education.
One of the interesting ideas proposed by Dr. Reddy was taking the help of AI to avoid future lockdowns. “During COVID-19, lockdown was introduced arbitrarily in India which turned into a major disaster for poor people,” he pointed out. “In the future, if you have an intelligent watch that measure all the vital statistics like temperature, heart rate, oxygen level, blood pressure and so on, then you can precisely say whether you have COVID and if you should self-quarantine. Then you may need a health passport...That type of technology is now possible.”
He pointed out that the cost of providing smart-watches to all citizens would be only about 20% of the loss to the GDP in 2020 due to COVID-19.
While acknowledging concerns such as job losses due to the advent of AI, he said that it will happen over a few decades which gives enough time to create societal mechanisms to deal with it through options. The options include Universal Basic Incomes, absorbing a higher number of people into more elastic jobs such as teaching, nursing, and so on.
Talking about misinformation, he proposed delaying social media postings by 24 hours during which AI technologies can be utilised to weed out deep fakes and malicious content.
The Hindu caught up with Dr. Reddy to understand more about his views.
What are the specific branches of AI that you currently find most interesting?
The most interesting area is machine learning. We have talked about machine learning in the past but we did not have the data and we did not have the computing power until recently. Now with the advent of those two things, the question is what can a computer learn by itself without your having to spoon feed it?
Alan Turing made a proposal in 1950 to make a computer learn like a child where it goes through the same experiences as a child and learns from that. He said, why don’t we teach the machine like we teach a child?
That idea has never been picked up until recently. We have been talking about it, but nobody knows how to do it. How do you put a computer in a classroom with all the other children and have it do the homework and submit the answers at the end of the day.
It will happen... The issue now is that it may not need to learn all of that because it already knows. The current systems have already absorbed knowledge in thousands of books.
The most interesting research problem is how machine learning can become an adjunct to human learning. What is happening with machine learning is that we are discovering new knowledge that we did not have before. It is not learning like human beings learn, but it creates new knowledge by discovering new patterns and solving new problems.
Governmental regulations are not able to keep up with the pace at which problems due to AI are arising. Do you think our ecosystems are mature enough to adapt to the pace at which technology is evolving?
The pace of change in technology is amazing. And societal systems that we have are not able to keep up with it. This will lead to a lot of disruption and dislocation.
We need to plan for it and if something becomes very serious, we have to do something. That is what happened with Australia. All the children were spending a lot of time looking at social network. So, they have now said no social network for people under the age of 16, which is one way of controlling it.
It may be impossible though because everybody has a smartphone. Parents and companies can control it, but there may still be ways to bypass it.
But that’s been the case. There are movies that come with the disclaimer that you have to be at least 12 years or older to watch them. But it’s not always controlled. So, we need to see what can be done.
With the help of AI, is there a possibility of prevention of pandemics in the first place?
No. Viruses are more powerful than human beings. There will always be pandemics. The only question is how to minimize the impact. We can’t shut down the whole society because of the pandemic.
What I’m saying is there will be another pandemic in the next 10, 20 years and there is no way to control it. But when it happens, we should not have to lock down everybody. We need to create some technology and AI enabled smart watches can help.
Are there particular set of problems that AI can very efficiently solve in a socially, economically and demographically complex country like India?
One is the language problem. If I take you from where you are 300 miles in a direction, you can’t talk to the people there. But with AI and speech technologies we may be able to overcome this. It’s a wonderful opportunity.
Do you think currently the benefits of AI outweigh the problems generated due to it?
My personal feeling is the benefits will be so big that the world will find ways to handle the problems.
AI can help with solving language differences, improving literacy and so many other problems. You can simply ask the computer to help you explain a certain idea to you. That’s a fantastic benefit to every person.
Do you think that way AI has the potential to democratize society?
Sure, provided you have a tool, and you know how to use it; And that means the right kind of education.
The opportunity is there. I don’t know whether society or when society will do it. It’s partly a function of what the government chooses to do. Even if the government wants to do it, you need to have teachers. So first we need to start to teach the teachers about AI.
What are your thoughts about the potential of AI to be a tool to exploit workers given the exploitative nature of many businesses in the name of efficiency?
Exploiting people has nothing to do with AI. If anything, AI ought to make it possible for people to work less but produce 10 times more. That might lead to some people being laid off. So, the question is, what do you do with them?
One way is that the government or the company must pay your salary partly. What we can do is maybe provide them with tax-intensive incentives where people are told only to work half the time and paid half the salary, and the other half comes from the government.
Published - December 09, 2024 08:00 am IST