Jason Jennings | Elevate your project management with AI. This guide for senior leaders explains how AI tools can enhance project performance through predictive foresight, cognitive collaboration, and portfolio intelligence. Unlock the potential of AI in your organisation and avoid the common pitfalls.

The Environmental Trade-off in Digital Infrastructure Development Digital development presents a double-edged sword. On the one hand, it boosts productivity through remote work, AI, and automation, with the potential to lift billions out of poverty. Yet, at the same time, the rapid growth of infrastructure required to support these developments will need a corresponding growth in decarbonisation to avoid a climate catastrophe. The German Advisory Council on Global Change highlights this contradiction: “uncontrolled digital change threatens to undermine the important foundations of our democracies” [1] . This article takes an in-depth look at how global institutions push the mantra of ‘digitisation’ as a developmental priority for nations while failing to adequately acknowledge the huge climate impact of this enterprise. This obscuring of consequences eases the way for a rapid extension of infrastructure that consumes billions of gallons of non-renewable resources annually. In this article, I suggest that detailed modelling and forecasting are one of the major pillars needed to address this dichotomy. I will set out an approach and resources for modelling the digital demand to design a more predictive approach to digital infrastructure builds. The Environmental Impact of a Data Explosion The amount of data flowing over global digital infrastructure has exploded 300-fold over the last 10 years [2] , with the next 20 years expected to see faster-paced growth on the back of the continued digitisation of life and entertainment, as well as from huge numbers of people in developing countries coming online for the first time. This explosion is a good thing—the UN’s Sustainable Development Goal (SDG) 9 aims to provide universal and affordable access to the internet by 2030 [3] . Access to the internet and digital services strongly correlates with improvements in education, healthcare and women’s empowerment. As increasing numbers of people come online, and the scale of their data use grows, a variety of digital infrastructure will need to be built or scaled up if the digital ambitions of countries and trading blocks are to be realised. Connectivity is one part of the solution—increased coverage of broadband, mobile and satellite will undoubtedly support these targets. But, ultimately, all that data traffic needs a destination point, in the form of data centres, which, unfortunately, require vast sums of power. In the USA, data centres are expected to consume 380TWh of electricity by 2027 [4] , almost 9% of the country’s total consumption. Ireland faces an even larger burden with digital infrastructure expected to consume 33% of the country’s total electricity by 2026 [5] , and potentially 70% of the country’s electricity by 2030 [6] . Ireland and the USA have reliable national power grids, but this is not necessarily the case in developing countries. In Nigeria, data centres and mobile towers rely heavily on diesel generators, burning nearly a billion litres of diesel annually. This is a country where the average annual mobile data traffic per subscription is only 6GB per year [7] , just over 0.1% of the average traffic from a UK subscriber. To achieve universal internet access for a population that is estimated to cross the 300 million threshold by 2036 will require an exponential growth in digital infrastructure. If Nigeria remained dependent on diesel generators, and data consumption on a per-person basis reaches the UK’s level of data traffic, then the country would consume 9 trillion litres of diesel a year—over 100 times the amount of diesel consumed by the entire world in 2022 [8] . This single event would create a climate catastrophe—even if the UK, France, Germany, Spain and the Nordics reduced their CO2 emissions to zero, this would offset less than half of this increase. This is of course the worst-case scenario. Grid infrastructure has developed across West Africa and there are a multitude of projects which are building green energy infrastructure. But there has yet to be a major MNO, TowerCo or data centre company which has shown significant year-on-year reductions in emissions. It is unjust to expect developing nations to slow down or halt their digitisation while developed countries reap the benefits of a digitised economy. Instead, alternative approaches to managing global emissions are needed. And this is where predictive analytics become a crucial tool for forecasting future demand. These tools and models will support the development of alternative strategies for power generation and implement methods to reduce emissions from digital infrastructure. A predictive tool that models national network traffic growth and compares it to projected digital infrastructure expansion will help identify underserved areas early, enabling better planning of digital and power infrastructure. Early planning allows for the integration of renewable energy, natural cooling solutions, and partnerships with sustainability experts to reduce emissions. Creating the Model: Traffic vs Digital Infrastructure To address these challenges, David Jones, an Associate of Cambridge Management Consulting, has developed a comprehensive model that examines global internet traffic on a country-by-country basis and compares it to existing and planned digital infrastructure within those countries. This model considers several factors: Population Growth: Increasing numbers of internet users Economic Growth: Rising wealth levels leading to more internet usage Internet Penetration: A growing proportion of each country’s population getting online Usage Patterns: Moving towards video transmission over the internet significantly increasing traffic B2B and M2M Traffic: Business-to-business and machine-to-machine Internet traffic growth This model projects internet traffic growth over the next 20 years, if data traffic growth follows a logarithmic curve, increasing at a decreasing rate. In Germany and other developed nations, the rate of traffic growth slows once it reaches a certain threshold, as there is a natural limit to how much HD video a person can consume. By comparing these projections with a database of over 10,000 data centres, including locations and power consumption, it is possible to identify regions with underdeveloped or overdeveloped digital infrastructure. Note: This model does not account for the growth in generative AI, which adds further demand on a strained digital infrastructure. For more information on this subject, see our recent article: Building an AI-ready infrastructure . Initial Results When we run this model and compare countries, what immediately becomes clear is the difference in scale between the growth of digital infrastructure and internet traffic. Ireland’s digital infrastructure is increasing at a rate faster than its internet traffic, while in countries like Bangladesh and Algeria internet usage is growing ten times faster than the digital infrastructure that supports it. David has modelled 76 countries and will be completing another 50 over the next few months. So far, the CAGR of internet traffic is around 30%, and the CAGR of data centres is around 12%. What’s clear from this graph is how the difference in growth rates compounds over time, and that as the years progress the gap between traffic and infrastructure widens. This shows that over time the availability of infrastructure will become a massive limiting factor to digital experience. Eventually, the lack of adequate infrastructure may even prevent citizens from accessing essential internet services.

In a previous article, Building AI-ready Infrastructure, we looked at the challenges that face the builders of digital infrastructure to create the massive engines that will power the ‘AI Revolution’ – in particular, the mega-data centres that will host the training systems used in Generative AI platforms like ChatGPT.  Most of the attention in the data centre industry is on these monsters, but there is more to it that we need to consider. This article looks at the other uses, applications, and implications of AI, and the infrastructure required to maintain them. The Growth of Industrial AI There are many flavours of AI, and although much of the current focus is on Generative AI, commercial applications use all sorts of other techniques to get the benefits that AI can offer. Indeed, there are some AI experts who think that too much emphasis is being given to the prominent large language models, and that the market will require a more diverse model for deploying infrastructure that will support real-world applications. There are many examples of industrial and manufacturing applications using AI already to optimise, for example, production-line efficiency in factories. These systems take data from sensors and devices (e.g. cameras), and then control the manufacturing processes in real time to improve efficiency, or to reduce the use of raw ingredients – a great example being the use of specialist glues in the automobile industry for sticking windscreens to car bodies – an AI platform has been in use to reduce the amount of glue used without compromising the efficacy of the bond. This may sound, trivial but the quantities used globally mean that even small proportional savings can amount to huge monetary savings. This type of application, used across multiple industries, has enormous potential for saving precious resources (or money), and many industries have been using these techniques for years. However, it is mostly the large manufacturers and processing companies that have been able to exploit this. Deploying this type of system can be expensive and usually entails situating a lot of processing power close to the production line. This excludes smaller enterprises from being able to take advantage as the barrier to entry is too high and involves maintaining IT kit that is expensive and difficult to look after.

Artificial Intelligence (AI) is the hottest topic in technology for many reasons, good and bad, but it’s happening and it’s here to stay, so how do we build the infrastructure necessary to support it? To start with, we should recognise that there are many forms of AI. The one that has created the most buzz is generative AI, as seen in ChatGPT, Meta's LLaMA, Claude, Google’s Gemini, and others. Generative AI relies on LLMs (Large Language Models) which have to be trained using vast amounts of data. These LLMs sit in data centres around the world, interconnected by vast fibre networks. The data centre industry has not stopped talking about AI for at least 18 months, as it gears up for an ‘explosion’ in demand for new capacity. Some of the most respected voices in technology have predicted immense amounts of growth in data centre requirements, with predictions of triple the current capacity within 10 years being at the conservative end. That’s three times the current global data centre market, which has taken 30 years or more to get to where it is today. And, when we say growth, we’re talking about power. AI systems will require three times more electricity than data centres currently consume. Depending on who you ask, that’s about 2-4% of today’s global electricity production. And we’re talking about tripling that, or more. Data Centres So, what is ‘AI-ready infrastructure’ and how are we going to build it? The two key elements are data centres (to house the AI systems) and networks (to connect them with the rest of the world). LLM training typically uses servers with GPUs (the chip of choice for AI) and, for various technical reasons, these work best when in close physical proximity to each other – in other words, GPUs work best in large numbers in large data centres. Not just that, but the new generations of GPUs work best in dense data centres, meaning that each rack or cabinet of AI kit needs a lot of power. Most data centres are designed to accommodate older kit that is not so power hungry. The average consumption globally is about 8kW per rack, although many still operate at about 2kW per rack. The latest nVidia (the leading GPU manufacturer) array needs a colossal 120kW per rack. The infrastructure inside a data centre designed for these beasts is complex: the cooling systems (GPUs run very hot) and electrical distribution systems are much harder to design and set up, and are also expensive. So, data centres for AI training systems are mostly going to be new, as adapting older facilities is a non-starter. So, where do you put them? Finding land next to the vast amounts of electricity required is increasingly difficult in many European countries, especially in the UK. Most of the utility grids in Europe are severely lacking in spare capacity, and building new grid connections and electricity generation is a slow and expensive process. The answer might be to locate these new AI data centres near new renewable energy generation sites, but those are few and far between, so land with access to power now carries a hefty premium. Small nuclear reactors could also be an answer but might take a few years to materialise – we know how to build them (witness the nuclear submarine industry) but getting planning permission to put them on land is another matter. All in all, the data centre industry seems to be at least a few years away from being able to provide the massive upgrade in capacity that is expected. Even solving the land/power problem leaves the issue of actually building a new scale of data centre, 10 or 20 times bigger than what most would consider to be a gigantic site today. It can be done, we can solve the engineering challenges, but these are huge construction projects. Networks What about the networks? Actually, although very little real research has been done on the impact of large-scale AI rollouts on existing networks, we might be in a better position. The fibre networks in the UK and many European countries have benefited from significant investment over the last few years, so coverage is a lot better than it used to be. That does not mean that fast and large fibre routes, which will be a necessity for most AI systems, are all there, but it will be easier to build out new capacity than it will be to find power. Still, what we really need is some serious research into the amount of data that will need to be moved about and how that maps with existing network infrastructure. All in all, we have more questions than answers. Some people in the infrastructure industry are sceptical that things will ever get to the scale that some are predicting, but most of us do expect it to happen – it’s just a matter of time, and the race has already begun. Cambridge Management Consulting Duncan Clubb is a Senior Partner at Cambridge Management Consulting, specialising in data centre and edge compute strategy. Duncan has extensive experience as an IT consultant and practitioner and has worked with many leading organisations in the financial, oil and gas, retail, and healthcare sectors. He is widely regarded as a leading expert and is a regular speaker at industry events. If you or your organisation require support preparing your Digital Infrastructure for the emerging AI-industry, you can read about our array of Data Centre services, and get in touch with Duncan Clubb, through our designated Telecoms, Media, and Technology service page.

This month we put the spotlight on Zoë Webster, Associate Consultant for AI, Digital & Innovation With over two decades in the Artificial Intelligence (AI) sphere, Zoë Webster is renowned as a practitioner and leader, recently recognised as one of AI Magazine’s Top 10 Women in AI in the UK and Europe (2024). At Cambridge Management Consulting, Zoë takes on the pivotal role of leading our AI initiatives and driving digital innovation. Leveraging her extensive experience in developing and applying novel AI techniques across diverse sectors such as retail, cyber security, defence, and health, Zoë is instrumental in shaping our AI strategy and implementation. Her unique ability to bridge the gap between the public and private sectors, coupled with her insights on the opportunities and risks of emerging technologies like Large Language Models, positions her perfectly to guide our clients through the complexities of digital transformation. Zoë’s expertise ensures that we remain at the forefront of AI advancements, delivering cutting-edge solutions that drive sustainable growth and innovation for our clients. An Introduction to Zoë's work Having been in the AI space for over 20 years, the past couple of years, since the launch of ChatGPT and the catapulting of AI into the public consciousness, have been in part eye-opening and in part déjà vu for me. The scale and reach are different to anything we have seen to date – I realised this when friends and family of all ages and backgrounds are talking about AI – but it is part of the well-cited technology hype pattern we have seen before in AI as specific techniques show promise (expert systems and neural networks, for example) and organisations see them as a way to solve current problems/challenges. I am fortunate in that I got into AI early. I describe myself as classically trained in that I learnt and experimented with the broad range of AI algorithms on different applications in my early career, so I understand that AI has much more to offer than whatever technique is currently in vogue. After developing and demonstrating novel AI techniques in a range of applications, I got the opportunity to learn more about the role of innovation to the wider economy and society through my time at Innovate UK, now part of UK Research and Innovation. From that, I understand the impact of technology and how business innovation can be accelerated given the right conditions and collaborations. My COVID-19 story includes the juggle of leading Innovate UK’s first COVID-19 innovation competition, to get critical grant funding out to businesses to ensure innovation could continue during this time, while attempting to home-school two children. During lockdown I joined BT, where I built and led their AI Centre of Enablement to scale up AI development and deployment across the company. Developing a machine learning model as a proof-of-concept is one thing, but it takes a whole other set of skills and approaches to successfully and safely deploy that model at scale and with real users, and then to repeat that for other models for different applications. Luckily, my breadth of experience as well as my deep AI expertise enabled me to set up and lead the team to specify and address dozens of AI opportunities. Even as the current developments in AI fail to quite live up to all the hype for everyone, organisations have an opportunity to apply the best and most relevant advancements to generate value, whether that is through customer acquisition, better customer service, better colleague experience, greater productivity or improved sustainability. This goes beyond the technology but to AI governance too, which means thinking carefully about how to practice AI responsibly. Working with Cambridge Management Consulting, I am excited to use my breadth and depth to help more organisations make the most of AI to create value in meaningful ways. To find out more about our AI, digital and innovation services, go to our Innovation service page or contact Zoë using the form below.