From the Highlands to the Cloud: Navigating data and AI

By Adele Fairclough

I was up in the Highlands recently for my summer holidays and as I was admiring the beautiful scenery, I realised something surprising – I had 4G. A few years ago, I was lucky to get a bar of phone signal. Now, I can read the news, stream video, and catch up on social media while at the most north-westerly point of the UK.

All this got me thinking about data, and the ubiquitous, but often unseen, part it plays in our modern lives. The word, plural of ‘datum’, comes from the Latin verb ‘dare’ – to give. I don’t think the ancient Romans knew then exactly how much ‘data’ would be ‘giving’ us now; nor how reliant we’d become.

Nowadays, organisations across industries and regions struggle with unwieldy amounts of data – across HR, marketing, finance and more. All of which must be stored, processed, moved, and analysed. And new technologies like artificial intelligence can help with some of this. But what’s really going on behind the scenes, and how can we manage such vast amounts of data efficiently?

Data and AI

It wouldn’t be a tech blog without mentioning AI. With recent strides in generative AI in particular, businesses across industries are racing to implement the technology, with many investing in their own tools as well as reaping the benefits of open-source platforms.

Many have been quick to espouse the various dangers and risks posed by AI – ranging from ‘it’s going to take our jobs’ to ‘it’s going to end the world as we know it’.

Fewer talk about the data that lies behind AI – the man behind the curtain. In order to run, AI models must first be trained on enormous amounts of data. And even generative AI – which claims to create ‘original’ content – in reality scrapes the internet for information (aka data) on which to base its creations.

Herein lies the real risk. AI is not (yet) some self-sufficient sentient being but relies on human input. And the data humans use can be corrupt, leading to ‘garbage in, garbage out’ results. In fact, of the organisations already leveraging AI, an average of five per cent of annual revenue was lost due to inaccurate or low-quality data.

AI outputs are only as reliable as their data inputs, so organisations must make robust data management a top priority. Unlike AI, data movement does not require human intervention. This means organisations can automate data movement to seamlessly and securely centralise disparate data sources from all corners of the organisation and create a single source of truth.

With this in place, organisations can conduct analysis using the freshest data and feed it to their AI models with the reassurance that outputs will be reliable.

The role of data centres

Another challenge with AI – and other compute-intensive technologies – is the sheer amount of power they consume. Naturally, this raises concerns for the environment as well as the impact on power grids. And these are valid fears, with major players like Microsoft and Google both citing increased data centre usage as the key factor behind their recent increases in emissions.

When it comes to where data is stored, most businesses think of cloud or on-premise. But there’s also a third option – colocation. The benefit of this approach is that organisations can be more mindful about where they locate their compute. For example, by choosing to locate their compute in colocation data centres in Nordic countries like Iceland, which have ready access to renewable energy, organisations can store and access their data without increasing their carbon footprint.

AI itself also has the potential to solve some environmental challenges. As the UN Environment Programme (UNEP) notes, AI can be used to improve the monitoring and measurement of methane emissions; to track air quality; and for measuring environmental footprints. Like any technology, AI is a tool. And we can use it to our advantage.

Is this the real life?

Data is so ubiquitous in our modern lives; it can be easy to forget it even exists. But in an information age, it runs the machine (both literally and figuratively). As we tackle logistical problems such as how and where to store data, and how to manage it as well as the energy it consumes, it’s important to keep sustainability front and centre. That way, we can keep enjoying the beauty of the physical world, as we enjoy the convenience of the digital.