Pradeep Khosla at AIMA’s 9th National Leadership Conclave
Collaboration among industry, academia and government is key to ensuring responsible AI innovation and ethical standards, said Mr Pradeep Khosla, Chancellor, University of California, San Diego at AIMA’s 9th National Leadership Conclave.
I’d like to start by saying that AI is like the whole evolution of humankind and society is all about understanding nature, capturing nature, and using nature to improve productivity. Because we as human beings are very limited people in terms of our physical capacity to do work.
If you look at steam power, electricity, mechanical amplification, everything we did was to create like amplification of our capacity to produce more effectively, agricultural revolution was the same thing. So, I think if you think about AI, it’s about expanding our capacity, our mental capacity, expanding our ability to think in ways that we have not done before. Before we were the brawn part of it, now we are the brain part of it. So that’s what I think about AI.
No, but as a productivity tool, it’s imperative in a world that is starved of growth, especially in certain areas of the developed world, for them to achieve future productivity gains is the only route for growth now. India is blessed with having a demography, which allows us to move up the per capita income chain.
If you look at all these revolutions, biotech revolution, agricultural revolution, IT revolution, at some point, all countries can play because the cost of technology becomes so cheap that everybody’s able to play.
If you think about education, it’s the only industry which has created the IT revolution, but the cost of delivering a unit of education has not gone down. In every industry, the cost of unit of whatever you’re selling has gone down, except in education. Our tuition goes up faster than inflation. The most recent article in the New York Times was that one year at a private institution costs $100,000. The median income in the US is only about $70,000 for a family of four. This says that most people in the US cannot afford to send their students to college, right? That’s how expensive it is.
But I think AI is going to change that. So let me give you one example, if you think about our education model, we have instructors, we have teaching assistants, we have lab instructors, we have a whole lot of support staff that are helping the student achieve the goals that they need to achieve. At UC San Diego, we are building an AI tutor. The idea here is that this tutor is trained on all the textbooks that your instructors use, knowledge of your instructor, and it’s available 24-7.
And we are running experiments right now with freshman courses in probability, in biology, in chemistry, where you have access to this instructor, and you can ask any question you want and this instructor now becomes your teaching assistant. The professor’s still there, but now the teaching assistant has a secondary role and most of the time it is this AI tutor that is your instructor. So that’s like one example of how it’s going to change. I can imagine that the professor might go away too one day. And we’ve been predicting that for 20 years. It has not happened yet. And I’ll talk about it later why it has not happened.
But at the end of the day, education’s going to become more outcome-oriented. The students will be able to customise their education experience. And institutions like mine will have to deliver that that ability to customise, not just a single fit everybody type of concept or notion for education, but the ability to customise your experience for every student.
Well, in this case, the teaching assistant may have all the answers as well. So, obviously, the form of evaluation must change. When Chat GPT came about, all the faculty members, just about every institution, were really very nervous about it, they were concerned about students not delivering original work. They were concerned about people going to the web and copying from the web and not delivering what they call plagiarism, original work, and I personally had a counterintuitive way of thinking about this. My view was it’s the combination of the student and Chat GPT that I’m going to hire.
I’m not going to hire just Chat GPT or the student. If this person can use this technology and coming up with the right answer, because Chat GPT is not perfect, right? Just like we are not perfect. The combination is more perfect than either one of them. It did not sit well with many people, but now I think all institutions are coming to this understanding that Chat GPT is a tool. We have to train our students how to use a tool very effectively and use it in ways to become a superior worker than they would be without chat GPT.
So, I would conclude that you still need a human engagement on this augmented basis of using a tool like artificial intelligence. Let’s look at the evolution of the human being. Our ability to think has not changed dramatically in the last, I don’t know, thousand years. Our ability to grow has not become faster in the last thousand years. We are a very slowly evolving organism. AI on the other hand is extremely fast evolving from knowing nothing to knowing.
Watch the full session- https://www.youtube.com/watch?v=NCPMHmXqc2g&t=145s
