By Amy Ji
Reuven Aronashvili, Founder and CEO of CYE Security, emphasized that “Quantum computing could soon transform large language model training…potentially breaking today’s cost barrier” (Forbes). Training our AI tools, such as ChatGPT, is extremely time-consuming and energy-intensive using classical computers (Forbes). According to Emilia Stuart from IQM Quantum Computers, unlike classical computers, which use bits, by using qubits, quantum computing can process many possibilities at once, drastically accelerating deep-learning training.
The “Hype”: Why the Excitement is Justified
Artificial intelligence is hungry for power. Training language models like ChatGPT and Google’s Gemini costs millions of dollars and is highly energy-intensive. Reuven Aronashvili, Founder and CEO of CYE Security, emphasized that “Quantum computing could soon transform large language model training” (Forbes) by tackling its biggest bottlenecks — speed, energy use, and complex optimization — making it one of the most exciting innovations.
Emilia Stuart from IQM Quantum Computers explains that, unlike classical computers, which use bits, quantum computing can process many possibilities simultaneously, drastically accelerating deep-learning training (IQM). Researchers from Quantinuum published a study with impressive findings, showing that their quantum computer consumes 30,000 times less energy than a classical supercomputer. Adding to that vision, Michael Epley, Chief Architect and Security Strategist at Red Hat, highlights the deeper fit between the two fields:
“There is a beautiful symmetry to the linear algebraic compute models that AI demands and that quantum computers can provide at massive scales, and as quantum information science matures and large scale quantum computers are made available via the cloud, we are on the precipice of delivering radical efficiencies for AI and transformal capability to tackle previously unreachable high-dimensional problems.”
Quantum computing can supercharge AI by rapidly testing millions of possible solutions simultaneously, enabling models to learn faster, use less energy, and find better answers than traditional computers can.
Why does it matter if quantum computing is making training AI faster and cost-efficient? Bernard Marr, from Forbes, underscores that the implications of quantum AI are profound and far-reaching, with the potential to revolutionize almost every industry. Megha Kalsi, cybersecurity Partner at AlixPartners, said at the Global AI Symposium, “When quantum and AI come together, it is going to cause this wonderful explosion in the cybersecurity industry, with cybersecurity tools and technologies that will allow us to identify threats faster.” In healthcare, Quantum AI can significantly accelerate drug discovery, enabling companies like Pfizer to bring new drugs to market more quickly. In finance, quantitative machine learning can be applied to perform risk analysis, optimize portfolios, and detect fraud much faster than classical computers or even the most powerful supercomputers (Medium)
The “Reality”: The Daunting Challenges We Face Today
Quantum computing is extremely powerful, in theory. However, today’s quantum computing hardware is not yet ready for AI.
The power of a quantum computer grows with its number of qubits, and today’s qubits are scarce and fragile. Currently, most machines have only a few hundred qubits, but AI-scale problems need thousands or even millions of reliable ones (Milvus). Researchers are also facing the problem of scalability with quantum computers. Qubits need to interact with each other to function, but maintaining connectivity is challenging because qubits are extremely fragile, and even the slightest interference can disrupt the field, resulting in errors and decoherence (TechTarget). To fix this, we need quantum error correction, which acts like a safety net, but it currently takes hundreds of physical qubits to create just one “logical” error-corrected qubit. That makes scaling up very difficult.
Despite today’s hardware limits, the foundation is strong. As Michael Epley observes: “Advances over two decades prove quantum computing’s mathematical foundations can be operationalized with cloud to serve the needs of AI and machine learning to realize the efficiencies needed to accelerate the AI revolution.” Once qubits become larger-scale, more stable, and error-resilient, Quantum AI will shift from a promising concept to a transformative force, reshaping industries from healthcare to cybersecurity and redefining the boundaries of what AI can achieve.
Amy Ji is pursuing her M.S. in Management of Technology at NYU Tandon, where she explores how innovation transforms the way we live and work.
