The Data (Use and Access) Bill is now officially a political football. Or ping-pong ball. Or whichever sporting sphere metaphor you prefer. On Monday, the House of Lords dealt a fourth blow to the government with a rejection of the latest amendment. Peers voted 242 to 116 to introduce greater measures for copyright protection. On Wednesday night, the House of Lords inflicted a fifth defeat on the government, voting 221 to 116 to amend the Bill so that AI companies must be transparent about the material they use to train their AI models. While some peers backed away from the amendment for fear of collapsing the Bill, the opposition remains considerable. Baroness Kidron has led the resistance to the Bill in the House of Lords, which has been heavily criticised by high-profile figures across the creative industries, and were joined this week by the director general of the BBC as well as the head of Sky. Insiders point to significant turning points since Labour took power as the root cause of the opposition, including AI Opportunities Action Plan author Matt Clifford arguing the UK should reform its copyright laws to attract AI investment, a government consultation describing an ‘opt-out’ method together with increased transparency obligation to AI firms as its ‘preferred option’, as well as Technology Secretary Peter Kyle meeting with big tech lobbyists, but not with creatives. Although one creative who doesn’t seem to mind AI is Björn from Abba, who told an audience at SXSW London this week that he uses it to help him write when he gets a creative block, even though it’s “very bad at lyrics”. Crikey, what are Oasis going to do then? Anyway, for a full round-up of where we are and what could be next, read our latest blog.
The Office for National Statistics (ONS) came under fire again this week after it revealed that its headline inflation rate for April was wrong. The UK stats body had been given the wrong road tax data by the Department for Transport, inflating the number of vehicles people were paying road tax on in the first year of registration. The correct rate of inflation should have been 3.4%, rather than the 3.5% figure given. The ONS has faced high levels of scrutiny recently, particularly its handling of labour market data, which it now has to reform after pressure from the Bank of England and Cabinet Office.
A new foot scanner that uses AI could detect heart failure in people on average 13 days before they would otherwise be admitted to hospital. The scanner takes 1,800 pictures per minute of the foot and lower leg, using AI to determine the level of fluid they contain. Water retention in the feet and ankles is one of the key warning signs that heart failure is developing into something more serious and potentially life-threatening. The scanner was developed by a Cambridge-based start-up, is roughly the size of a smart speaker, and can be mounted on a wall. Elsewhere, Foresight, an NHS England plan to train AI on patient data (previously reported here) has been paused. The British Medical Association and the Royal College of General Practitioners said they were not aware that GP data was being used to train the model, asked NHS England to refer itself to the Information Commissioner, and demanded a pause in the processing of data.
Join us next Monday 9 June 13:00-17:00 BST at the UK National Data Library Symposium, hosted by King’s College London. Our CEO Louise Burke and Director of Research Elena Simperl will take part in panels alongside many others to focus on the transformative role of data in research, policymaking, and innovation. Talks will delve into how the National Data Library could improve data infrastructure and systems, how AI could enhance data utility, and much more. Find out more and register for the event.
And finally…a team at ETH Zurich has developed an AI-driven robot that is trained to play badminton. The four-legged (and one-raqueted) robot can follow the trajectory of a shuttlecock and hit the projectile back, with a considerable amount of agility and accuracy. While the team behind it believes it could pave the way for advanced, autonomous systems, I reckon it could pick up ping-pong pretty quickly. That’s just what the Data Bill match needs. An AI robot to sort it out…not remotely a terrible idea.
David and Jo
PS: It’s the last chance to book Solid World on Monday 9 June - a project working to give the individual more control over their data. Book your ticket here.
PPS: One of our members, Ovation Data, published a white paper this week about clean, structured data. Check it out!
