Tech innovator Rapid Innovation posits that combining quantum computing with artificial intelligence (AI) constitutes one of our most exciting technological frontiers. The article "Quantum AI: Pioneering the Future of Innovation in 2024" explains how traditional classical computers are already reaching their limits as AI technologies rapidly evolve.
"This is where quantum computing comes into play. Quantum computing leverages the principles of quantum mechanics to perform computations at speeds and scales that are unattainable by classical computers," it explains.
The synergy between quantum computing and AI can potentially revolutionise various industries. Quantum-enhancedAI can significantly accelerate machine learning tasks, making them more powerful and efficient, and paving the way for a brighter future.
In a talk with FutureCIO, tech leaders and experts discussed the potential of quantum-enhanced AI, its limitations, and its potential impact on AI technologies.
Future of possibilities
For Alan Priestley, a VP analyst in the Emerging Technology and Trends Research team at Gartner, quantum computing is currently an immature technology, ten years away from showing commercial advantage. He adds that it is only applicable to specific use cases and cannot handle as many algorithms as classical computing can. However, he acknowledged that it could solve some of the problems that AI technologies face today.
Charlie Dai, VP and principal analyst at Forrester, agreed that the technology's overall readiness is still far from commercial-scale adoption.
"While the research and development around quantum machine learning (ML) remain active, the transformational impact of generative AI driven by foundation models has largely offset the investment in other domains, including quantum computing," Dai said.
As with many novel computing technologies, the question isn't just " can it do the job?" The question is "can it do the job better?
Joseph Yang
"As with many novel computing technologies, the question isn't just " can it do the job?" The question is "can it do the job better?" said Joseph Yang, the general manager, HPC & AI, of APAC, and India at Hewlett Packard Enterprise.
"Information Technology companies and their research partners are in the initial phase where they are trying to develop effective tools to analyse problems and estimate the quantum and classical computational resources that will be required to design a hybrid solution better than today's best practices," Yang said adding that the quantum computation applicability to conventional AI and ML algorithms is still an area of active academic research.
Real-world use cases
Forrester's Dai said the experimentation of quantum AI is primarily in combinatorial optimisation. In the finance sector, use cases include risk modelling and optimisations of trading strategy, asset pricing, and portfolio.
"For healthcare, it includes optimising radiotherapy treatments, generating targeted cancer drug therapies, and creating protein models. And for energy, cases include energy exploration, seismic survey optimisation, reserve and spot trading optimisation, and reservoir optimisation," he adds.
Yang posits that real-world applications of quantum-enhanced AI are still emerging in Asia, but significant investment is being focused on developing the broader quantum ecosystem.
"Despite the theoretical yet unproven potential, quantum-enhanced AI presents transformative opportunities across various industries globally. Enhanced cybersecurity with quantum key distribution (QKD), accelerated drug development through molecular simulations, improved financial modelling and risk analysis, faster climate simulations for better environmental policies, and optimised logistics processes are all being explored," he added.
Challenges of quantum-enhanced AI
Gartner's Priestley said that building large-scale quantum computers is one major challenge of quantum-enhanced AI.
"Conventional wisdom says we probably need around a million qubits to create a system with a business advantage versus classical. There are different approaches, and different companies will look at them differently," he said.
Further, Priestley considers it an engineering challenge to construct a system that can hold millions of qubits and complex dilution refrigeration systems to cool large-scale quantum computers.
Developing efficient quantum algorithms for AI tasks remains a hurdle, while noise, error rates, and qubit stability remain critical challenges to building reliable and scalable quantum computers.
Charlie Dai
"From a technical perspective, developing efficient quantum algorithms for AI tasks remains a hurdle, while noise, error rates, and qubit stability remain critical challenges to building reliable and scalable quantum computers,” said Dai.
"In addition, while hybrid approaches that combine classical and quantum processing are being explored, bridging classical and quantum systems seamlessly is very complex," he added.
He adds that a significant shortage of quantum experts and governance remains challenging for organisations.
Yang explained that integrating quantum technologies will challenge IT organisations at every level. Organisations should consider factors such as "bringing a dilution refrigerator holding qubits at a fraction of a degree above absolute zero into the data centre to learning how to craft probabilistic quantum algorithms."
Advice for tech leaders
For Priestley, organisations can start evaluating quantum computing for specific tasks depending on their needs.
"Businesses today need to be looking at whether they can leverage this with their business or whether they should start evaluating how they can do it today. And given this current market state, they need to build skills and use the capabilities that companies offer," he said.
Tech leaders should evaluate these questions: Does this potentially help me in my business? How can it help me in my business? What skill sets can I build to leverage this technology and experiment with what's available in the market? How can I use this? Where does it fit into my business? What do I have in my business that would need this?
AI is still further out in time. Maybe it will come, but the current AI algorithmic approach for AI is not suited for quantum computing.
Alan Priestley
"But AI is still further out in time. Maybe it will come, but the current AI algorithmic approach for AI is not suited for quantum computing," posits Priestley.
Yang said, "Quantum-enhanced AI will likely remain an active area of research for the foreseeable future. Time will tell if quantum will provide an advantage in conventional AI algorithms, if new quantum native approaches will be required, or if it will remain a stronghold of conventional computing. To create a resilient enterprise roadmap in the face of this uncertainty will be a challenge in itself, but enterprises can be confident in the first steps."
He added that the first few steps include employee training focusing on clarity in basic quantum concepts to help the organisation evaluate the future potential of quantum-enhanced AI and raise awareness of quantum secure communications and quantum attacks on conventional cryptography.
"Companies with unified and collaborative management that work with all their stakeholders through transparent communication are better placed to see success in a quantum-dependent future," he said.
He underscored the importance of beginning the discussions around quantum-safe and post-quantum cryptography (PQC), a type of cryptographic algorithm that is secure against quantum computers, to better prepare for quantum adoption.
"Beginning the CISO conversation will prepare the enterprise to critically evaluate their potential quantum-enhanced AI future," Yang said.
