How to control the new gold rush in Scotland? Ray Perman takes the long view of disruptive technology – history can point the way to a strategy that brings real benefits. And it leads to Shetland.

According to the Financial Times there are at least 24 planning applications currently for “hyperscale” data centres in Scotland – those that would consume more than 100 megawatts of power. They are causing consternation and anxiety in the places where they are proposed.

It’s easy to understand the objections. They would be very large – one west of Edinburgh would occupy 30 hectares, or 42 football pitches – cause considerable environmental damage, drain huge amounts of electricity and water, but provide few permanent jobs and little benefit for the communities in which they would be based.

History tells us that not all, if any, will be built. Disruptive new technologies tend to provoke mania and AI (artificial intelligence) is following a similar path to the railway mania of the 1840s, North Sea oil in the 1970s and the dot-com boom of 2000. Hyperbolic prospectuses describing phenomenal future benefits will persuade investors to pump unthinkable sums into companies which have no track record.

Many will lose their shirts

AI will bring some lasting benefits, just as the railways, oil and the internet have done. Some companies will survive and prosper – and a handful of individuals will become very rich. But many of the exaggerated claims will never be achieved, many firms will go bust and many investors will lose their shirts. It has always been thus and there is no reason to think AI will be different.

In 1845 at the height of the Scottish railway mania, speculators had secured parliamentary permission to build 3,000 miles of track north of the border. Less than ten per cent of it was built. In the 1970s millions were poured into developing fabrication sites around the coasts to build oil drilling and production platforms. One – Portavadie in Argyll, built with Government money – failed to win a single contract. Most of the others got one, some built several, but none managed to establish a lasting business.

In the late 1990s dozens of young companies raised significant amounts of capital on their promises to enable farmers to buy fertiliser online, or to build a network of the Scots diaspora, or to help you book flights and hotels, or to bet on sports games. They were all either bought out. Or went bust.

Home grown hopefuls

The companies proposing to build data centres in Scotland are not the titans of the AI world – Alphabet, Meta, Microsoft, Open AI and Anthropic, all of whom are investing hundreds of billions of dollars in the US. The companies active in Scotland are more home grown and hoping to catch some crumbs from the rich men’s tables.

Apatura is a York-based property developer which has turned to data centres and battery storage sites as a promising new stream of opportunities. It has secured planning permission for a 24-megawatt battery storage centre near Glasgow and is proposing to build data centres west of Edinburgh and near Falkirk.

ILI (Intelligent Land Investments) began by developing eco-friendly residential developments before diversifying into green energy. It is a Scottish company with some well-kent faces on its board, including former Lord Provost of Glasgow Michael Kelly and former UK minister Brian Wilson.

With an interesting marketing twist, it has announced plans for three new hyperscale data centres across Scotland, each named after an ancient Roman Stoic philosopher: Fife (Cato), East Ayrshire (Rufus) and North Lanarkshire (Aurelius). The proposed Cato centre near Auchtertool, Fife, would cover an area equivalent to almost 40 football pitches and be up to 35 metres high, roughly the height of a 12-storey block of flats.

It is easy to understand how communities and even councils can feel that they will be branded as luddites, standing in the way of progress by trying to block these developments. But again, history can give us pointers to how we should react.

Sullom Voe shows the way to go

a levy on every barrel of oil landed went into a fund to benefit the community long after the oil boom had passed. Today that fund is worth £475 million.

In 1974 Shetland was besieged by oil companies desperate to bring their oil ashore and seeking to build terminals across the islands. There was a national interest in allowing development, but the islands wanted to control it. The council, led by its chief executive Ian Clark, and the local MP Jo Grimond secured a private Act of Parliament, the Zetland Act, which gave the council the ability to compel oil companies to share a single terminal at Sullom Voe.

It also imposed a levy on every barrel of oil landed, which went into a fund to benefit the community long after the oil boom had passed. Today that fund is worth £475 million.

Wind farm developments can provoke similar opposition to data centres. The community on the Isle of Gigha decided to develop their own, with the revenue benefiting the whole island. In the village of Fintry the community agreed to support a large commercial wind development provided it could own one of the turbines. Its revenue has financed community projects for the past 19 years.

The Scottish Government has admitted that its policy on data centres which was formulated in 2021 and was positive towards new developments, has not kept up with the number of proposals and the size and impact of the hyperscale centres. Just the aggregate energy demand of the current proposals calls for a national rethink.

But in the national interest we should not forget the local. Communities need to feel they have a stake in those developments which do go ahead.

Further reading

APRS AI Data Centres Campaign Scotland’s Countryside Charity

Data Centres: SPICe Spotlight

Dancing Ladies turn 30: Isle of Gigha