AI
should not be considered just a
computer science
subject anymore, instead it should be taught to students and professionals in every field, says Arnab Bhattacharya, professor of computer science and
engineering
at IIT Kanpur.
Arnab, who did his PhD in computer science from the University of California in Santa Barbara, says Gen AI is everywhere now and that all educational fields must change in response to it.
“Just like we teach
Programming
101 to everybody, AI 101 should be taught to everybody. Everyone need not know the mathematical details, but what AI can and cannot do should be known by everyone now. It is very, very important.”
Arnab points out that AI is already being used extensively in most engineering fields, and engineers in crucial fields like civil, mechanical, and chemical engineering need to understand that using AI systems that are black boxes can have serious consequences.
“What’s going on inside the AI system? What is the integrity of the data? How much bias is there? Can you believe the output? How much can you believe? How much is explainable? These are the things that a person who is applying AI, or influencing some policy on AI, should know.” He says there’s no threat of job losses from AI in the immediate future, because right now the world is riding the rising wave of AI. “But once it stabilises and saturates, then AI will take over more and more jobs, even skilled jobs,” he says.
However, this period of flux that AI is ushering in will also elevate certain job profiles and skills, and create many that didn’t exist previously. “Take for example data. AI is very much data dependent. So, understanding the domain of data is extremely important. If you want to understand data, if you want to know which features to look at, which features to engineer, which features to avoid, what is going on in the data, all of that requires deep domain knowledge, that is a skill that needs to be developed,” he says.
WHAT TO STUDY
Computer science students looking to specialise in AI, Arnab says, should do it deeply. “When I say deeply, I mean both in terms of the technical aspects – like what goes on inside neural networks, how to program them, what exactly fine-tuning means, what a tokenizer is exactly supposed to do, etc – and also the social science part of it.”
The social science part includes areas like AI bias, AI ethics, fairness, etc, all of which are expected to have an outsized effect on the field.
General computer science students, Arnab says, need not focus on the technical aspects of AI but should still not ignore the social science aspects of it. “They need to know how AI can affect things.” They should also focus, he says, on being proficient in programming languages and ensuring mediocrity is kept at bay, because that leaves them susceptible to being replaced by AI.
English (US) ·