The whole world has been obsessing over AI for over 3 years now. But what’s the next big thing? Or the one after that? Here, we explore three emerging technologies that should be on everyone’s radar, including one that you’ve probably never heard of.
It’s easy to look at how artificial intelligence is discussed and think that that’s it. There is no other technology that will ever matter. But if technological innovation has taught us anything over the last 150 years, it’s that it rarely stands still. Just when you think you’ve got the hang of it, something comes along and – bang – the landscape shifts again.
While it is perhaps foolish to spend too long staring into a crystal ball when it comes to tech, there are reasonable signs of what the next three major technologies will be that are set to transform the world, and more specifically, mark further shifts in the future of the accountancy profession. Those are: digital twins, quantum, and the one that you almost certainly won’t have heard of, neuromorphic computing.
Digital Twins
Is this new technology? Technically, no. But can we still call it emerging? Most definitely. While the world has been distracted by AI, the technology underpinning digital twins has quietly been gaining traction.
A digital twin is a virtual representation of a physical asset. That can be anything from a simple object right up to an entire organisation or city. Unlike a simulation model, digital twins are designed to faithfully maintain synchronicity with their real-world counterparts through tight integration and live data feeds. They typically require robust data management and governance and, for larger systems, significant computing power. By precisely creating a digital version of a physical item, it becomes possible to more accurately model and predict future outcomes, test decisions and monitor performance.
Often, digital twins are talked about alongside blockchain technology and tokenisation – indeed, ICAEW referred to them in our response to the recent FCA consultation on tokenisation – however the two are distinct. There are opportunities for digital twins in the world of blockchain and tokenisation, effectively enabling the tokenisation of real-world assets, an approach that is starting to develop in some supply chains to track inventory and provenance of physical items in near real-time, or allowing the representation of physical investments and their changes in value in digital form.
Digital twins have huge potential in the accounting profession. As well as supply chain modelling and tracking, they can support oversight of manufacturing operations, map out process optimisation opportunities, and pave the way towards more continuous auditing techniques, as well as supporting the planning of audits by using digital twin technology to simulate and stress-test key audit risks. And as finance teams take on more responsibility for sustainability reporting, digital twins can help to model ESG performance, making the capturing of sustainability data less burdensome.
Quantum
It’s fair to say that most people don’t realise quite how far quantum computing has developed over the last few years. Far from being science fiction, the technology very much exists with the large technology providers like IBM and Microsoft already starting to tout their quantum solutions. But what actually is quantum computing?
Unlike traditional computing, which works on the basis of a series of 0s and 1s in alternating states, quantum computing uses a branch of mechanics where two states can exist simultaneously. The unit of information used in quantum computing, called a qubit, can be a 0, 1 or both, and the principle of quantum entanglement means that qubits can interact with and impact each other instantly. This opens up vast computational power with the ability to explore many solutions simultaneously – a bit like trying every route through a maze at once.
Quantum computing is not without challenges. Qubits are, by their nature, unstable and difficult to measure. This places significant demands on hardware – extreme cooling close to absolute zero, shielding from environmental noise, and careful management of quantum error correction, a critical focus of current research.
There is also the cryptography challenge. If – or indeed, when – quantum computing becomes commercially viable, classic encryption techniques will become obsolete. This is a significant risk to our modern interconnected lives.
However, scientists are starting to solve these problems, and as quantum computing gains traction, so its potential use cases start to emerge. For accountants, these use cases are likely to be some way off yet, but will include sophisticated risk and scenario modelling, anomaly detection, business planning and optimisation and advanced forecasting.
Neuromorphic Computing
Here’s where we get into serious crystal ball territory. Neuromorphic computing is a term used to describe a completely different way of working – more radical than AI or even quantum.
Ultimately, computer architecture hasn’t changed dramatically in decades. They use transistors – electrical switches that are either ‘on’ or ‘off’. Even quantum computing still works on the same basic principle of two states, albeit simultaneously. In terms of power consumption, the approach is inefficient – billions of transistors are needed in an average computer, all requiring power, and while qubits can do more with less, they require significant cooling and infrastructure to be relied upon.
Brains, on the other hand, work very differently. They use neurons, each of which is connected to thousands of other neurons via synapses. Neurons don’t work on a binary basis, instead responding to variations in signals from any combination of the neurons they are connected to. Equally, if a neuron is not being used, it stays idle. This enables substantially more capacity to process information, with substantially lower energy needs. It’s estimated that the human brain uses only the same energy as a 20-watt light bulb. Neuromorphic computing, therefore, is exploring the potential to replicate the behaviour of neurons and mimic brain functions, with a view to delivering substantially more powerful compute, with a fraction of the energy requirements.
Given the focus on AI’s current levels of energy and resource consumption, the need for a better solution is accelerating. Neuromorphic computing has the potential to be that solution. In addition, neuromorphic systems are, by their very nature, more flexible and adaptable than traditional algorithmic systems. While generative AI capabilities have been designed to closely replicate human logic, in their current architecture they cannot and will never truly think the way humans do. This is the potential of neuromorphic computing, and it will transform all aspects of society.
There is, however, one small problem with the development of neuromorphic computing: understanding how brains work. While the field of neuroscience has grown rapidly in recent decades, scientists still have a long way to go before they truly understand the mechanics of brain activity. The good news is that neuromorphic computing is not strictly the result of fully understanding the brain. It is an attempt to capture a handful of the brain's most useful computing principles, particularly its ability to process information in parallel while using very little energy. On this basis, the next tech revolution might not be as far away as we think.