AI and data

Artificial intelligence and data-enabled technologies have huge potential to transform our lives and society as a whole. Ensuring they are used safely and effectively is essential for the UK’s wellbeing, security and economic growth.

The power of data and AI

Almost every aspect of our lives generates or uses data. From the phones in our pockets and the entertainment we stream online to the energy we consume in our homes, the data we help to produce is growing exponentially. So too is the ability to analyse it.

When used effectively, data can provide insights that can be used to improve healthcare, spot disease outbreaks or discover new treatments. It can be used to make journeys on our roads more efficient and can help manage critical infrastructure such as the energy network.

The vast data sets now available have also helped with the development of a form of artificial intelligence known as machine learning. These computer systems learn from data, examples and through experience rather than having to be pre-programmed to carry out complex tasks.

Many of us now interact with a form of artificial intelligence every day through search engines, social media and voice recognition software. As the field develops, it is likely to percolate into our lives in ever more surprising ways, so it will be important to build new governance structures to ensure it is used fairly and transparently.

Data for society

The era of “big data” – where vast amounts of information can be gathered, stored and analysed – makes it possible to see patterns or trends that would otherwise be missed. During the Covid-19 pandemic, for example, anonymised mobile phone data helped public health officials to model how the virus might move across international borders and within populations. The use of large scale data could also help to improve the response to humanitarian disasters or help monitor for and give warnings about natural disasters such as earthquakes.

Data is also being used to develop digital technologies that can help us. Training AI-powered image recognition on medical scans and digital images has yielded results that suggest these could help with the diagnosis of a number of diseases. Image recognition, speech-to-text transcription, virtual assistants and obstacle navigation can help people with vision, hearing, mobility and learning difficulties.

Any use of data, however, comes with challenges relating to privacy and security. Everyone has the right for their data to be stored safely and used responsibly. It is important to ensure flaws and biases within the data don’t accentuate inequalities in society. Failing to do so could undermine public confidence in the technologies that use data.

The Royal Society's work on data and digital technologies has been investigating the potential opportunities and barriers of these new applications.

AI for scientific research

The scientific method relies upon collecting and analysing data. With techniques such as machine learning, examining large amounts of data and identifying patterns within it can be done at greater speeds than ever before. They can also offer insights and solve problems that have previously been elusive.

AI can help astronomers, for example, identify distant objects such as pulsars from the noise of background signals created here on Earth. Machine learning can help to identify changes in satellite images that reveal deforestation or help to predict the structure of proteins from DNA sequences, speeding up drug discovery.

In some cases, AI tools can produce answers or solutions that may seem unconventional to our eyes. The use of these technologies has the potential to disrupt the way science is done in the future. This means it is also important to ask how they can be used most effectively and in ways that researchers can interpret and understand the results they produce.

The Royal Society’s report, Science in the age of AI, details how artificial intelligence is changing the scientific research endeavour with expert insights and recommendations for policymakers.

Digital technology for the planet

To limit future climate change, we need to find ways of rapidly reducing carbon emissions that result from human activity. With the help of data and AI, it may be possible to find more efficient and sustainable ways of going about our lives and doing business. Intelligent systems can help us to reduce the carbon footprint of buildings and manage energy networks to make the most of renewable energy.

The enormous power of machine learning to analyse data is also making it possible to keep a closer eye on the changes happening on our planet. Image recognition systems can identify changes to landscapes and biodiversity, and monitor greenhouse gas emissions from different sources. AI-enabled computer modelling can also improve weather forecasts and predictions of extreme events such as storms.

But data storage and the computing power needed to run AI systems require energy themselves and can lead to carbon emissions, so it is important to find ways of reducing their impact on the planet too.

In its report on digital technologies and the planet, the Royal Society outlined a vision for how AI and data-enabled systems might be mobilised to tackle climate change in the future.