As the world has been taken over by AI and ML, science is now chasing the countless benefits that quantum computing will offer in forecasting weather, the market and possibly life. Quantum physicist, Dr Rukhsan-ul-Haq, who works with IBM talks about his journey into quantum computing and details how life will change in the near future
KASHMIR LIFE (KL): At the peak of Covid19, when science was desperate to understand the virus, why we did not use the Quantum Computer and why we relied only on classical and supercomputers?
DrRUKHSAN UL HAQ (DRUH): Quantum computing was used in understanding the virus, especially in drug discovery, where the classical and Super-computers have limitations.
Unlike the classical understanding that biological molecules are made of atoms, deep-down research has shown that even the atom itself is a quantum specie. Quantum physics gives the structure of an atom and shows its actual composition.
Biological molecules are also made up of atoms or we can say quantum species, so in order to understand this quantum computing is used. Classical computing gives. an approximation because of its limitations but quantum computing has no such issues. The fact is that quantum computers gave us insights and an understanding of the virus and helped in understanding its mathematical epidemiology – how it spreads and creates epidemics).
However, quantum computing is in the infancy stage. These became public only in 2016. Quantum computers have Qubits as a basic structure of processing. The first IBM Quantum Computer had only five Qubits. At present, we have crossed a hundred Qubits at IBM. SimilarlyGoogle have 50-53 Qubits in its quantum computer.
KL: Since you are the first Kashmiri quantum physicist who worked at IBM, the organization where even Nobel laureates work. Tell us about your journey to IBM.
DRUK: When I was in 10+2, Physics and Mathematics fascinated me. In 2000, I joined Amar Singh College for undergrad admission. I decided to study Physics, Chemistry and Mathematics. It got complicated as I had not studied Mathematics at the secondary level because in was a biology student. Somehow, I succeeded in convincing my teachers and that is how Physics and Mathematics became central to my study. At that time, there were no proper facilities. Thanks to the library of Amar Singh College which helped me manage my studies better. Teachers were supportive even though they were not able to technically help me. They gave me a free hand, especially my mentor Assadullah Sahab from Baramulla. Though he was teaching somewhere else, still, he helped me in understanding basic Physics.
Later, I was admitted to the University of Kashmir for my master’s degree. Quantum Physics (Quantum Mechanics 1 and Quantum Mechanics 2) intrigued me. I think my entire journey to IBM started right there. When I got to know about the mathematical beauty and conceptual structure, it became clear to me that I have to go into Quantum Physics only.
Around 2011, I went to Bangalore to join the Indian Institute of Sciences (IISC) Bangalore, where I got admitted to JNCASR (Jawaharlal Nehru Centre for Advanced Scientific Research). I was fortunate that I got Prof NS Vidhyadhiraja as my supervisor. I started working with him on quantum and condensed matter physics. My PhD spanned over 5-6 years. We explored more about Quantum Computing.
During my PhD, I started exploring Topological Quantum Computing, on which Microsoft has been doing more research.
I collaborated with two leading quantum physicists – Prof Loius Kauffman, from Chicago University (founder of Kauffman Polynomial) and a French quantum expert, Michel Planat, which brought me close to Quantum Computing. We were working on Decoherence, that is, preserving the Quantum Information. Unlike Classical computing, Quantum computing information is more fragile. Classical computing information is easy to preserve in hard drives, but for Quantum information we have to create methods for its safety and integrity, a process called topological protection.
KL: You devised a method for the protection of Quantum information during your research. Is it in vogue?
DRUH: Yes, the methods we devised are used at two levels. First, whenever we create any Qubit of quantum information, we have to protect that so the information does not get coupled to some system. The information should not be corrupted, because it will lead to wrong calculations. Our research was about Error Correction, an engineering mechanism of putting information in Qubits without getting the information loss and how to correct the errors if any.
My second contribution was in the magic state that is creating such a chain of Qubits that cannot be simulated on any classical computer in the world. Quantum communities around the world adopted both of our models.
Another of our contributions is mentioned as a separate chapter in the textbook of IBM. We gave the mathematical model of a Qubit, which is fitted inside the superconductors (rather than semi-conductors in classical computers). In order to understand the existence of Qubit inside the super-conductors, there is a method called Schrieffer-Wolf Transformation. We made a mathematical method in order to understand the Qubit, which was widely recognized. When IBM did its first seminar on Quantum Technology in 2020, in which our method was showcased and now it is being taught in renowned universities across the world.
KL: Is this method mentioned by your name in the IBM textbook?
DRUH: There is actually a reference to my paper in the textbook and a detailed explanation of what that method is.
KL: When did you complete your PhD, and when was the first Quantum computer made public?
DRUH: I completed my PhD in 2017. It was in 2016 actually when the first Quantum Computer was launched. I was doing PhD at that time and I was among the few people who registered for the IBM Quantum Experience. By now millions of people use that. It is ago to place when it comes to Quantum Computing.
KL: Moving from Wadwan village in Budgam to IBM? What was the experience?
DRUH: It was a kind of privilege for a person like me to get an opportunity to work with IBM. Until my PhD, I was more in academics and research, but I always wanted to become a scientist and it was realized in some sense. It was like a dream come true.
KL: Normally young people in Jammu and Kashmir look for teaching positions in colleges and universities after completing their PhD. But you choose a different career?
DRUH: My family also suggested that I should look for a teaching job. I wanted to remain abreast of new research frontiers. The fact is there were developments taking place when I was busy with my PhD. Scientists started using AI (Artificial Intelligence), ML (Machine Learning) and Deep Learning technologies.
I had two options at that time – either to stay at the conventional level and miss the opportunity or get into the newer things.
ML has been there since 1930. It fascinated me. I wanted to explore the combination of ML and Quantum Computing. I also wanted to be in the industry rather than academics. I always wanted to work in companies like IBM, Google, and Microsoft. Initially, I did not directly join the Quantum computing industry. I first got into ML because there were many of my friends in this field who shared their experiences and opportunities.
By 2017, ML was a buzzword. People were experiencing the impact of it and celebrating this technology. Before joining the industry, I had done a brief Post Doc in China.
My first post-PhD project was with Johnson and Johnson in Bangalore. I was given the role of a Lead Data Scientist. There, I along with my team developed a forecasting platform to help the supply-chain management right from manufacturing to the end user. They had a conventional forecasting platform but that was inefficient. We developed an AI-based platform that had much more accuracy in determining demand and supply. We got appreciation from the top management for the model we built. This gives me confidence that I can work in the industry.
After that, a life-changing incident happened when IBM approached me to work for their Bangalore-based centre. They were looking for a person with a PhD degree in Quantum Physics, who also has the industry experience and a good background in Quantum computing and AI. For me, it was an amazing opportunity.
KL: What is your present role in IBM Quantum?
DRUH: I have a technical role as a Quantum scientist, which means I have to look for new developments in Quantum computing, implement them, and make them more accessible. My job is to make quantum machine learning models for predictions and forecasting in fields like biology, drug discovery, and disease outbreaks and make them run on hardware Quantum computing devices. One of the most important domains we are working on the IBM is Material Simulation – to make the materials more efficient.
Richard Feynman, a renowned physicist at the 1982 conference of IBM and MIT, made it very clear that Quantum materials of any nature like materials in batteries, green energies, superconductors, chemicals, and drugs, can be simulated or understood only through Quantum Machines. Although at that time, there was no such machine but it motivated many to build Quantum machines. Today we have a Quantum machine and we are working on the material simulations.
When we build ML models, typically they are more complex and need more computing and processing power to run. Presently we have three levels of processors – CPU (Central Processing Unit), GPU (Graphical Processing Unit), and TPUs (Tensor Processing Unit). Deep Learning/ML models can only be run on TPUs. TPUs are expensive and are not environmentally friendly. That is where the domain of Quantum AI comes. It makes sure that Quantum Computing can be used to make AI models better. Thus Quantum models will be more efficient in terms of computing, consume fewer resources, work faster and are simple to use.
In classical AI or ML models, a CT scan or an X-ray is passed through the ideal model to detect the abnormality. Most of your data or news is summarised through ML models. However, these models have limitations. Quantum NLP (Natural Language Processing) is a Quantum approach to language processing. It can do stuff beyond classical modelling. It helps us to make models better and more efficient. It gives us the models that help us understand climate change.
KL: How does the Quantum computer look like and is it different from the classical computers that we use?
DRUH: Quantum Computers will not make classical computers obsolete. Both will go hand in hand because classical computers have their own strength. Quantum Computers will be used in those fields where classical computing is struggling.
Quantum computers are very big in size and almost fit in a big room. We can access the Quantum computer without being physically around, through cloud computing. We just need to create an account on the cloud-computing platform, which will open the interface on our classical computer. By clicking on the interface, the operation will be made in the Quantum computer, very far from your location. In future, they may also shed their weight and mass and become like classical computers.
KL: Our computers have fans to cool down the hard drive. But Quantum Computers require below zero temperature in order to function efficiently. How it is being done? Does this make the Quantum Computing system more complicated?
DRUH: Quantum computers must be kept at very low temperatures to maintain Coherence in quantum information, which is fragile. Thus, Cryogenic Systems of absolute zero is maintained for the Quantum Computers. It is an engineering challenge to maintain such a low temperature but we are able to do it through different thermodynamic processes.
Other major challenges in Quantum computing include the prevention of information corruption, noise cancellation, error correction, and upgrading of Qubits.
KL: With AI and ML taking over a lot of jobs, are there new jobs that are being created? What major changes did the last two industrial revolutions bring in the job market?
DRUH: Humans are prone to error and defect but machines bring high precision, are fast and manufacture automatically. We just need to set the machines correctly and they will do the job. The aim is not to replace human effort, but to make it more efficient. If a machine can do a certain kind of job then why should we relegate it to a human being? Humans need to focus on doing creative jobs.
KL: How should we look at the challenge of machines taking our jobs?
DRUH: On one side job market gets affected adversely because of this technology but on the other hand many other jobs were introduced like Quantum AI scientist, Data scientist etc. Many years before no one knew of data science but today it is a booming industry. Everyone wants to be a data scientist because it is a very sophisticated, challenging and interesting job. Data science is the most appealing job in India and the hottest job in the world.
Quantum computing will have the same effect. I think these revolutions are creating more and new exciting job opportunities. Al is at its maturing phase but the Quantum computing has a long way to go.
We are currently living in the information age where skill and speed matter and humans cannot compete on that level. We have to confess the fact that machines can do better and more precise operations than humans in some domains if programmed correctly.
KL: With AI playing everything, humans hardly remember data. Is there a stage in the offing when AI will acquire emotion and threaten human beings?
DRUH: Some people do talk about how this technology will start running upon us. However, I see it differently. For me, AI is the next level of software development. Initially, Software development was based on instructions. AI is also Software but has a mechanism of learning in it. It has learning behaviour. It has the potential to correct the operation if it was wrongly done in the past. We do anticipate that, as technology evolves, it will become more intelligent than what we have. But I think the button will always be in our hands. We can always have that control. The decision of how a machine should operate will always lie with human beings as is true with any technology.
(Mujtaba Hussain processed the interview)
