IN DEPTH
AI Without Data? The National Data Library Must be AI-Ready From Day One
Emma Thwaites, Director of Global Policy and Corporate Affairs at the Open Data Institute (ODI) was part of this year’s Rewired conference. Here she looks at the how a National Data Library could work and what is needed for it to be a success.
THE idea of a National Data Library (NDL) emerged in the Labour Party Manifesto ahead of this year’s General Election. Since then, there’s been ongoing discussion about its shape and scope.
My organisation, the Open Data Institute, has contributed thoughts on the NDL to the AI Action Plan (
https://tinyurl.com/odiNDL1), reviewed research and written about the concept more broadly (
https://tinyurl.com/odiNDL2). The original vision from Labour for the NDL was to “bring together existing research programmes and help deliver data-driven public services, whilst maintaining strong safeguards and ensuring all of the public benefit”. This will only be possible if we create a well-thought-through piece of infrastructure. This involves ensuring access to high-value datasets, building infrastructure that the public can trust and ensuring that people have the right skills to use and maintain it. Many of the things that information professionals think about daily, no doubt. As we’ve argued at the ODI, without data, there is no AI. As a country, we need well- structured and well-governed data to support AI stacks. To illustrate just how critical this point is, one study found that analysts typically spend 80 per cent of their time (
https://tinyurl.com/ odiNDL3) preparing data for AI use. To ensure the same challenge doesn’t blight the NDL, we must design it to be AI-ready from the outset.
October-November 2024
Emma Thwaites is Director of Global Policy and Corporate Affairs at the Open Data Institute.
The NDL will require high-quality datasets curated from existing public sector bodies, research organisations, and beyond. However, data often sits in legacy IT systems that don’t “talk” to each other and vary enormously in format and quality. For example, a great deal of data still only exists in PDF documents stored in rudimentary databases. With these kinds of challenges, it’s tempting to try to design a comprehensive data architecture from the get-go, with accompanying detailed rules and reporting requirements. But it’s important not to stifle good practices where they exist or, indeed, to make the task ridiculously daunting for those who don’t already have high-quality data. It’s also critical to understand the cost of digital (and data) transformation when public finances are under pressure. While
INFORMATION PROFESSIONAL 27
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52 |
Page 53 |
Page 54 |
Page 55 |
Page 56 |
Page 57 |
Page 58 |
Page 59 |
Page 60
