“Moore’s Law revisited – where does quantum computing take us?*”

Robin Fieth, BSA Chief Executive, explains how building societies and credit unions can learn from the development of quantum computing.

Article first appeared in Society Matters

Robin Fieth, BSA Chief Executive explains how building societies and credit unions can learn from the development of quantum computing.

This article first appeared in Society Matters

Looking backwards, looking forwards.  Or is it the other way round?  As we look forward to gathering once more as a whole sector for the 2022 Building Societies Conference in Liverpool during the first week of May, I was reflecting back to our 2018 Conference in Manchester, where I spoke about themes for today’s boards – and asked whether Moore’s Law is the biggest strategic question of our age?  Moore’s Law, the observation that the number of transistors in a dense integrated circuit doubles every two years, has been with us since Gordon Moore, the co-founder of Intel, set it out in 1965.  Conventional thinking is that exponential curves top out.  But the question of where are we on the Moore’s Law curve still remains valid, as does the question about what innovations will overlay new factors and new curves that continue the exponential trend in processing power and capability?

The likely answer to the second of these questions is now becoming clearer, and to a novice like me, quantum computing doesn’t seem so much about overlaying a new exponential curve, as shifting to a whole new dimension.  At its simplest level, it’s quite difficult to comprehend why the ability to get a simple logic gate to process both zero and one simultaneously is quite so revolutionary.  Until you try to visualise it!

Functioning commercial quantum computers may still be a way off.  Qubits (quantum bits) may be able to hold the values one and zero at the same time, but they are also inherently unstable.  Vibration, temperature and other environmental factors can all cause them to lose their quantum-mechanical properties, resulting in errors.  Huge resources are being brought to focusing on solving these problems.  And perhaps sooner than we might expect.  So it is not too soon for boards to start the learning process, understanding the potential opportunities and threats, and formulating early responses.

A recent Harvard Business Review[2] article might help start those conversations in five areas: simulation; linear systems; optimisation; unstructured search; and factoring and encryption.  Looking at each of these in turn from our sector perspective:

Simulation: in its simplest form (quite scary just writing that), the ability replicate the very best manual underwriting – and significantly improve on it.

Linear Systems: taking existing machine learning and neural network developments to new dimensions in product recommendation systems, say.

Optimisation algorithms: determine which decision in a given scenario is most likely to produce the desired outcome.  Imagine a whole new range of individually tailored dynamic mortgage products which automatically adjust to borrowers’ constantly changing circumstances.

Unstructured Search: solving the needle in the haystack problem of finding information in large unstructured databases.  Early attempts to use social media activity as a component of credit scoring may be seen as pioneering rather than bizarre.

Factoring and Encryption: perhaps the most obviously concerning.  Quantum computing could crack the prime factorization encryption that underpins much of current global internet security and privacy infrastructure.  That means just about everything of interest to state-sponsored hackers and cyber criminals.

Quantum computing – definitely one to have on your radar!

Next steps: You can follow Robin on twitter @bsaceo


* [1] Cramming More Components onto Integrated Circuits, Gordon E Moore (1965), Electronics

[2] Quantum Computing for Business Leaders, Jonathan Ruane, Andrew McAfee and William D Oliver (2022) Harvard Business Review