Dr Vishal Rao cautioned that AI cannot be used in everything because it is not a replacement for natural stupidity.
Published Aug 15, 2024 | 8:00 AM ⚊ Updated Nov 26, 2024 | 2:40 PM
Dr Vishal Rao (in left) at Dakshin Healthcare Summit
Artificial Intelligence (AI) is everywhere – it’s nearly impossible to go a day without hearing or reading about it. But what role does AI play in surgery, and what can we expect in the future?
During the inaugural Dakshin Healthcare Summit in Hyderabad, Dr Vishal Rao, Country Director of Head & Neck Surgical Oncology at HCG Cancer Centre in Bangalore, explained that when we talk about AI in a broader sense, there are four stages: Narrow AI, Limited Memory AI, Theory of Mind AI, and Self-Aware AI. Currently, we are in the stage of Limited Memory AI.
“For example, just a couple of months ago, we were in Sunnyvale, California, at the world’s largest robotic surgery summit. It was a proud moment representing India there, and it was fascinating to see how machines are interfacing with humans in reality. We even saw a driverless car and a robot that came to our hotel room to take our order. What we are seeing now is the third stage—Limited Memory AI—with driverless cars. Elon Musk has already made strides in this area,” explained Dr Vishal Rao.
He said that however, what Stephen Hawking warned about is the potential danger when we reach the stage of self-aware AI. This is when AI achieves autonomous consciousness and emotive intelligence, and it’s at this point that society might need to start worrying.
“Back in October 2019, during a talk, I predicted that in 10 years, we’d be on the verge of cracking the Theory of Mind in AI. We all have to accept that we are moving at a rapid pace. Fast forward to 22 June 2022—I received an email stating that Blake Lemoine had been put on gardening leave by Google because he had successfully cracked the Theory of Mind. This was just two years after the COVID-19 pandemic began,” he said.
He then said, “Now, think about this: if we’ve already touched on the Theory of Mind in AI just a year or two ago, we’re witnessing an unprecedented acceleration of artificial intelligence, far beyond our current understanding.”
“As a robotic surgeon, I’m learning to play Xbox video games because you need to be really good at that. So, I sit with my kid, playing Xbox and PlayStation, asking for help with hand controls because I need those skills for surgery. In the South, we perform some of the largest head and neck surgeries using robotics. We’ve even invented our own instruments and processes. For me, it’s been a tremendous learning experience. Robotics is here to accelerate, accentuate, and assist human healing—not to replace humans, at least not right now,” Dr Rao said.
Now, when you combine robotics with artificial intelligence, machine learning, and the way these technologies are evolving, it’s truly exciting.
“We’re seeing how AI is adapting, learning, and starting to give us more suggestions and even take on some independent tasks. This development is particularly thrilling because it spans across diagnostics, therapeutics, and theranostics—covering a wide range of medical areas. We’re on the brink of AI and robotics merging together in a very meaningful way,” Dr Rao said.
However, he said he would like to add a word of caution. He said he often tells junior enthusiasts who are eager to use AI in everything: AI is not a replacement for natural stupidity.
“The idea that AI is the solution for everything is misguided. So, we need to be careful and deliberate in how we integrate AI into our practices, rather than assuming it can solve every problem,” he added.
He said that he had the privilege of securing three patents in India—three new drugs, all targeting cancer. The process has been incredibly enlightening, particularly in how AI has played a role in this journey. One of these drugs is already on the market, while another is in the pipeline. The third drug, for which a patent has been granted, heavily utilised AI in its development.
“We partnered with a group in Connecticut, which is one of the few organisations that employs large language processing alongside evidence mining. This is an advanced system that not only understands conversational AI but also comprehends protein interactions and discussions,” said Dr Vishal.
He said that one of the most impressive applications of this AI was during the preparation of a US FDA application, which is typically a very cumbersome and expensive process.
“At that point, I’d already completed the preclinical phase but wasn’t certain about the additional 20-30 tests I might need. Instead of spending hundreds of thousands of dollars on trial and error, we fed all our data into this AI system. Now, it’s important to note that pharma companies are usually reluctant to share their proprietary data, so gathering information is often a challenge,” he said.
However, the AI was able to crowdsource and analyse the data, giving me incredibly detailed insights within just five days. It even created a complete US FDA submission for me, predicting outcomes and simplifying the entire process.
“In a country like India, where drug discovery is often discouraged in favour of generic drug production, this experience was a game-changer. It showed me that AI has the potential to revolutionise drug discovery, making complex processes more efficient and accessible,” Dr Rao added.
(Edited by Neena)
(South First is now on WhatsApp and Telegram)
