Photo by — @siimonfairhurst

With Our Heads In the Clouds

A Reflection on Data Centres and AI Alongside National Strategies

At most time when I read national strategies pertaining to artificial intelligence (AI) there are many that do not mention data centres. To a great extent it is hard to talk of AI without talking about data, and the storage is often not thought of or planned for in strategies. After all AI is not a metaphysical construct much like any other digital project. If we are to talk of AI and impact, or AI for good then we have to talk about data centres.

There has been an attempt to map global data centres, although I find it highly unlikely that this is a comprehensive map:

Recently there was an article in Analytics Insight arguing that the data centre industry was going to increase greatly due to the rise in cloud demand.

“The main difference between the cloud vs. data centre is that a data centre refers to on-premise hardware while the cloud refers to off-premise computing. The cloud stores the data in the public cloud, while a data centre stores the data on company’s own hardware. Many businesses are turning to the cloud. In fact, Gartner, Inc. predicted that the worldwide public cloud services market has grown to 17.5 percent in 2019 to total US$214.3 billion. For many businesses, utilizing the cloud makes sense. While, in many other cases, having an in-house data centre is a better option. Often, maintaining an in-house data centre is expensive, but it can be beneficial to be in total control of computing environment.”

So yes, it is a question of earning and cost. However there is increasingly a question of security and sovereignty, wanting to own the digital infrastructure in question, such as proposed in the recent data strategy by the European Union.

Is there however an appropriate amount of planning on how to make the material cost of data centre cheaper, or reduce the use or consumption of data? There seems to be of no great concern of companies such as Facebook, Google, and the other large technology companies to have people reduce the amount of data that is used. Of course this speaks against the consumption of goods, their revenue model is advertising and so on.

Yet can we dare to think that there are no consequences for our usage of cloud platforms? I think at least we should ensure that the places we use or operate are as responsible as possible. This does of course relate to reducing carbon emissions, ensuring labour rights, and being as responsible in every way by the manner in which we use technology.

Data is there in a physical location at a data centre, and we must not forget this when we talk of artificial intelligence — as much as we (sometimes) think about where a banana is grown when it arrives in the store. Fair trading practices seems strange in the context of digital, there are no FairTrade servers, yet maybe there should be. We are removed from a large part of the supply chain, yet it is not impossible to conceive of ideas to help make it more transparent or responsible than it is in its current state.

This is #500daysofAI and you are reading article 269. I am writing one new article about or related to artificial intelligence every day for 500 days. My current focus for 100 days 200–300 is national and international strategies for artificial intelligence.

AI Policy and Ethics at Student at University of Copenhagen MSc in Social Data Science. All views are my own.

AI Policy and Ethics at Student at University of Copenhagen MSc in Social Data Science. All views are my own.