The American economist’s words have been brought into sharp focus by the UK’s labor productivity stagnation. According to the Office of National Statistics (ONS), UK productivity was 13% below the G7 average in 2020¹

If these numbers seem abstract, figures provided by the Resolution Foundation think tank brings home the stark reality². The UK’s productivity gap means the average worker is £11,000 worse off today than they would have been if we had continued to increase productivity on-trend. It's a problem that matters to us all.

Finding a solution is far from easy. As every business leader knows, there are four ways to increase productivity:

• Reduce your cost base (for example, materials, or energy costs)
• Increase your capital investment to improve efficiency, capacity or create new capabilities
• Increase the size of your labor force, to increase specialization, efficiency, or scale output
• Generate ideas and apply them

The problem is that the first three items on the list are finite. Demography, for example, might not quite be destiny – nations with smart strategies or mineral or other natural resources have grown rich and powerful – but it sets limits on macroeconomic growth. The application of ideas does not.

The UK’s Innovation Strategy suggests 51% of the UK’s economic growth was driven by innovation (between 2000 and 2008)³. Drill into the sources and you find that much of that was not just new technologies, but process and business model innovation.

It is our only realistic way to close the productivity gap. It is as important for our absolute prosperity - as individuals, as a society, and as a nation, in comparison with previous generations - as it is for our relative power in the international system. This is why it is at the heart of the UK’s Science Superpower ambitions, laid out in what Minister George Freeman has taken to describing as our goal to be ‘an innovation nation’.

Automation, robotics, and artificial intelligence (AI) reduce nations’ dependence on demography to create wealth and to scale their armed forces to deter or fight. They are fundamental to our future.

As businesses, we have seen this effect. Far fewer workers are now needed to generate $1m of revenue by companies listed on the S&P500 than ever before⁴ .

But while we are producing immense value with fewer people, overall, our productivity is lagging. This perhaps partly explains the growing inequality gap, which is harmful to society, to democracy, and to the individuals involved.

It is both a growing challenge and an opportunity. Training AI systems is becoming cheaper and cheaper⁵. AI has been everywhere for a number of years – seen in internet browsers, curated social media feeds, map apps and more - but as the Alan Turing Institute points out, ChatGPT is the first innovation to feel like AI.

Wharton Business School Professor Ethan Mollick reports productivity gains ranging from 30% to 80% (varying by industry) from the application of ChatGPT to tasks in the workplace. Contrast that with our productivity gap as a nation of 13%.

In 1997, Professor Hans Moravec of the Robotics Institute of Carnegie Mellon University wrote his paper ‘When will computer hardware exceed the human brain⁶. In the study, he described the landscape of human competence, where the rising tide of AI capabilities was slowly subsuming the land which once represented those things we considered to be cognitive abilities unique to humans.

In an edition of Nature in 2021, a paper described how chip floor planning, the design of semiconductors, was now being undertaken by AI. By 2023, most, if not all, of the major semiconductor designers and manufacturers were employing AI to undertake the task.

In December last year, Anthropic reported that ‘we have moved further away from reliance on human supervision, and close to the possibility of a self-supervised approach to alignment⁷.

The answer to the question Moravec posed in 1997 - namely ‘When will computer hardware exceed the human brain?’ – written 26 years ago, was: ‘the 2020s.’

This is far from the end of the road, though. Scaling laws – applying more computational power, to more data, with more parameters - lead to predictable increases in performance: we can expect things like ChatGPT to continue to get better and better linearly.

However, as we increase the scale of our models, we start to unlock transformational ‘phase shifts’. Foundation models went from not being able to translate human and machine languages, for example, to doing so, in an instant. There was nothing to indicate they would gain this capability, until suddenly, with increased computational power, they had it.

We are entering a world that will usher in a revolution in how we interact with information. This change will be more profound than the introduction of the printing press, or the internet. Many of the assumptions around how our society and systems function, of what matters and how the world works, will be upended.

We need to be clear-eyed in the profundity of the challenge we face, the opportunity, and the risks. We need to understand that innovation is critical to our future prosperity and security. And we need to remember that the translation of the technological ‘phase shifts’ into process and business model innovations will drive this revolution.

That is our mission, whether in business, government or any other sector. And we’ll only succeed if we work together to achieve it.

Keith Dear is Managing Director of Fujitsu’s Centre for Cognitive and Advanced Technologies

Fujitsu’s Centre for Cognitive and Advanced Technologies brings together innovation in areas including digital twin technology, quantum computing, and artificial intelligence, as part of Fujitsu’s commitment to developing a high-skills UK economy. Providing a direct link to our work in Japan, the Centre provides a focal point for collaboration between industry, Government and academia, as well as enhancing the UK’s ability to draw on Japan’s leadership in areas such as supercomputing.

To learn more about how technology and the ideas in this article could enable your organization to address productivity challenges, contact The Centre for Cognitive and Advanced Technologies

---

¹ Office for National Statistics H/T @richardaljones, Twitter, 18/3/23

² Resolution Foundation. 2023. 15 years of economic stagnation has left workers across Britain with an £11,000 a year lost wages gap. Press Release. 20 March 2023.

³ BEIS. 2021. UK innovation strategy (publishing.service.gov.uk), BIS. 2014. Innovation Report 2014 (publishing.service.gov.uk), NESTA. 2013. Innovation Index 2012 | Nesta

⁴Bradley Saacks, Bank of America data, via Twitter: https://twitter.com/Jason/status/1620582399573884928?s=20.

⁵ The rise of Skynet - by Miguel - Genuine Impact Newsletter (substack.com)

⁶ Moravec, H., 1998. When will computer hardware match the human brain. Journal of evolution and technology, 1(1), p.10

⁷ Bai, Y., Kadavath, S., Kundu, S., Askell, A., Kernion, J., Jones, A., Chen, A., Goldie, A., Mirhoseini, A., McKinnon, C. and Chen, C., 2022. Constitutional AI: Harmlessness from AI Feedback. arXiv preprint arXiv:2212.08073

Topics:
Digital transformation  Life at Fujitsu  Responsible Business  Quantum   AI   Sustainability  
Industries:
Financial Services    Public Sector    Defence    Retail    Energy    Manufacturing   Healthcare    Transport   Education