The International Year of Quantum Science and Technology is officially underway, following the UNESCO inauguration this week. It marks 100 years since the birth of quantum mechanics – as well as an inflection point in quantum computing and other related technologies achieving real-world applications.
We are seeing significant sums of money invested in quantum computing, alongside huge financial bets on AI, with nations competing to gain commercial and strategic advantage. ChatGPT surpassing 300 million weekly users, and the vast stock market fluctuations after DeepSeek’s AI launch, underscore just how far the adoption and normalisation of AI has come in the past 18 months. So, will the same soon be true of quantum, and how can businesses start unlocking quantum value?
Quantum vs. AI
It’s worth a quick jump into the differences between a quantum computer and AI. AI is the latest evolution of silicon-based computation. The technology performs increasingly advanced maths and statistics that can help predict and model events.
Generative AI is essentially an incredibly capable prediction engine for what we are likely to expect to see, hear, or read based on a prompt. AI requires a large data set to train on, and the training entails high power computation, but it can then run rather quickly.
Quantum computers, on the other hand, are fundamentally different as they make use of different physics.
This results in a large amount of parallel processing – combining several operations in a single step. At the moment, quantum hardware lags behind the hardware used for AI, because it requires a lot of development to keep it stable and create machines at large scales. For example, for a quantum calculation, you ideally need to isolate the quantum “bit” from pretty much everything else, or there is a risk it will do the maths wrong. A quantum calculation doesn’t necessarily require vast amounts of data because you can “just” set it up to solve a maths problem, but typically there is at least some data somewhere.
Pros and Cons
This difference in operating principle means the two technologies are good at different things and have pros and cons relative to one another. They can also work together – particularly when looking forward to a more mature quantum computer.
In both cases, organisations will spend billions on developing and connecting new hardware, creating new algorithms, and making use of new products that consume and generate data in enormous quantities, from a wider range of sensors and data sources, to solve problems that are currently beyond reach.
This will require new skills and techniques that haven’t been fully invented yet. And in both cases, the resultant tools will be accessible to a vast audience across multiple sectors, probably through the cloud.
Limited boardroom engagement
But despite this degree of overlap, there is a stark difference in boardroom engagement between quantum and AI. There are a few reasons for this: some cultural, some technical. First, quantum tech is hard to explain.
You inevitably end up discussing qubits, entanglement, Hamiltonians, and a variety of other complex technical terms that aren’t relevant to business applications. This is a failure of communication and a reflection of how we train scientists, where it’s often not necessary to understand the benefit.
As a result, quantum tech is typically seen as far away, wildly expensive, and extremely complex – in other words, the province of the scientist with a white coat. Whilst there are use cases that are a long way off, some are more accessible near term, and most people will use quantum via cloud hardware rather than owning a quantum computer.
The narrative on AI has moved from The Terminator and The Matrix style science fiction to how it can help users solve their day-to-day needs – and with that, an articulation of near-term value. The same could be true of quantum in the next three to five years.
Unlocking the value of quantum
We are already starting to see the convergence of quantum and classical tools. For example, through its CUDA-Q platforms and partnerships with start-ups like ORCA Computing, NVIDIA is building hybrid devices that work at the intersection of quantum and classical systems.
Similarly, Google is talking about quantum AI, and users can already integrate quantum services into apps using Amazon’s Braket service. A more mature quantum ecosystem – like the existing AI market – will probably contain very few companies that make the hardware, a few more that make low level software and run data centres, and a lot that base their products and services on it.
Ignoring quantum as an emergent technology is an error, as it will deliver market value.
Quantum computing offers huge opportunities to solve problems exponentially faster, simulate molecular structures to accelerate drug discovery or design new chemicals, speed up training of AI models, improve weather modelling, and more. The use cases with the greatest near-term benefits are where we need support in making complex decisions at speed, such as in financial portfolio management or supply chain optimisation. The business case for these is easier to calculate and explain. To many users, there may appear to be little actual change; just the system becoming more capable.
Taking the quantum leap
The quantum hardware is not yet ready to be truly competitive in the aforementioned applications yet, as current systems are a combination of slow, small, unreliable or unstable. But this will be solved – and companies need to be ready to run when the quantum starting bells rings. As with AI, no one will want to be last.
The near-term to-do list for companies is to understand where there is benefit to quantum tech. It has the potential to be better at some things, and worse at others. Businesses should start building the capability to use quantum computers, through periodic benchmarking, testing, and trialling. This means targeting use cases with near-term business value and benchmark what you can get against what you need to unlock returns. It may be that something quantum-inspired gets you most of the way there today – such as for maintenance scheduling and supply chain management.
It’s also important to build the skills base for quantum tech. The skills that allow us to exploit AI and data – mathematics, problem-solving, an ability to spot business value from technology and communicate it – are exactly the same as those required for quantum compute. It’s a different language, with some nuances of course, but to ignore one is to ignore elements of the other. As with AI, there is also a need to be mindful of arising risks. Look no further than the National Institute of Standards and Technology’s recent release of post-quantum cryptography standards for that – these standards highlight the need for organisations to be prepared for a quantum-enabled “hacker”.
Unlocking the benefits without succumbing to the hype
It’s important that organisations strike a balance between recognising the benefits of quantum and getting entangled in hype. Quantum compute is an evolution of cloud compute with, as ever, new capabilities and trade-offs – it should be part of a trade space when thinking about a high-performance compute roadmap, but the sensible users will pick their spots and use the technology accordingly. Quantum will not replace AI. AI will not stymie quantum. Instead, they will be mutually supporting tools in a broadened “toolkit”.
We’ve been here before – we had “big data”, then machine learning, and then AI.
We will at some point have quantum and AI, then something else on top of that. In the meantime, organisations should assess the threats and benefits of both quantum and AI; understand where, when and how high-performance computation, regardless of platform, can deliver business benefits; and ensure they have access to the skills they need to make use of them.
Because when the starting gun is fired, it will be a race.
- Digital Strategy
