Artificial intelligence (AI) may have made several strides but it still needs a ‘catalyst’ to realise its complete potential. As always, every technology has its limitations — in this case, there’s only as much that classical computers can do with computational capabilities. That’s where quantum computing steps in to boost the action of AI so that more complex issues can be dealt with.  

Not many have dabbled in this field, unlike L Venkata Subramaniam, who is the Quantum Distinguished Ambassador at IBM. His focus has largely been on AI Reasoning and Quantum AI/ML. His vision is to set up Quantum Computing as a leading area of research in India, and his efforts have been truly in the right direction.  

His journey into this field is peppered with interesting nuggets, which he shares with Jibu Elias — Content and Research Head of INDIAai, as part of a freewheeling chat. Read on to know more about his foray into the field of AI, the challenges he faced, and how India can gain a competitive edge with quantum computing.  

The tryst with AI 

Venkata’s journey is far from conventional — he had completed his Ph.D. two decades ago in telecommunications, which largely dealt with coding and information theory. Previously, he pursued a degree in electronics engineering. It was a stroke of luck that IBM had opened up an office on the IIT Delhi campus, where he was studying. He gave it a shot and despite getting through the interview, he wasn’t convinced.  

“I questioned the director about why they wanted to hire me, since I had a telecom background. That’s when he told me that in the 1970s, those who had invented speech recognition at IBM had similar academic qualifications,” described Venkata, adding that it was early on that he had his first brush with AI; it was referred to in a different way back then. 

Speaking about the field of quantum AI, he calls it more of “progression”, since there were certain gaps in AI that had to be addressed. These were computational complexities that had to be resolved. Venkata shared that although these are tough to tackle, quantum really has the potential to provide an advantage.  

The potential of quantum computing  

 It was in the 90s that Shor's algorithm and Grover's algorithm proved that one could get exponential gains. Back then, there was no certainty that anything could happen in reality, until the quantum devices were rolled out. That’s when the growth of the hardware was visible.  

“That’s when you start believing that the hardware is going to be able to run some of these algorithms, so it is worthwhile to develop them,” he added.  

Shedding light on the current status, particularly at IBM, Venkata said that we are in the “Noisy intermediate Quantum” era. Interestingly, they have been able to integrate some of the classical cloud infrastructures, and quantum, to bring them together seamlessly. Further, since cloud technology is advancing fast, there are also ways to reduce the latency. 

“In quantum, there are issues like fidelity, errors, the number of qubits is still small, etc. But at IBM, we released the 100 Plus last year, which is our last quantum device that is part of the system one architecture. This year, it's going to be the second generation. There were certain things in the way that qubits were connected that did not allow scaling in the first generation architecture,” he explained, adding that in the newer generation architecture, it will be easier to  

make better interconnects between the qubits. This will allow the drawing of more power from those qubits and at the same time, reduce the interference.  

Navigating the challenges 

At IBM, quantum has been isolated into four parts. At the bottom is the hardware layer, for which there is a requirement for physicists, and those who have in-depth knowledge of quantum mechanics. It also needs. Another level higher are people that are the bridge between hardware and software, who know a bit of physics but also know all about the software principles.  

“These are the algorithms people are people who need to have an understanding of how superposition and entanglement work. Say I'm a courier company and I want to deliver all these shipments throughout Delhi, but I want to save on the fuel costs, the number of trucks and drivers, and other issues, but that can’t be solved on a classical computer. They can be dealt with on a quantum computer, but you need to understand how to bring in superposition and other things,” shared Venkata.  

At the top layer are developers — here, they do not even need to know about superposition; instead, they are provided with libraries on optimization, machine learning, chemistry, and everything else. They just bring in their data and problem and run it using these. Unlike earlier, top companies like IBM do not necessarily want people to be good at physics, okay, not everybody is good at physics; there are various tools that can help you stay at the top layer.  

How India can gain a competitive advantage in quantum 

“IBM has been also propounding the message to build an ecosystem. We make sure to bring universities, industry startups, the government, and everybody else together; that's what we're trying in India also. Also, my goal, especially as the leader for quantum in India is to ensure that we kickstart the ecosystem, because, without it, it will remain as it is, and when the students will learn, they will move to other countries,” concluded Venkata.