Data: the Lifeblood of UK AI Startups

Today we have published a new paper detailing how startups – and by extension the UK’s innovation economy – can maximise the opportunity of the National Data Library.

We know that data is one of the three legs of the AI stool alongside compute and talent. Without it the whole thing falls over. The UK, with its deep public sector data pools should be well-positioned to exploit this advantage but as things stand datasets are often locked behind impenetrable gates, outdated and unreadable.

The NDL could change this. As work continues while Ministers and Parliamentarians are away, we propose the NDL is built with a firm focus on:

Supporting Departments and ALBs clean up their datasets so that they are fit for use ensuring that, as far as possible, datasets are standardised, findable and complete.

Establish clear mandates and incentives to drive open data. This should include strengthening enforcement of open data principles, introducing a full TDM exemption and introducing secondary publishing rights, following the lead of many of our European neighbours.

Building mechanisms to create all new singing and dancing datasets to support a mission-led government. Without a unified mission data initiatives like these can flounder. The NDL should act as the research arm of the mission boards – corralling the data needed from right across government to deliver on their priorities.

Time and again we have seen enormous breakthroughs emerge from experimentation and analysis of data. AI advances make this even more possible, something clearly recognised by Government in documents like it’s Life Science Sector Plan. However, without access to vast datasets these breakthroughs will remain buried.