Nuffield’s Ada Lovelace Institute last week produced a damning summary of the exuberance around digital contact tracing.
“Inaccurate, irresponsible or imprecise use of data or technology can undermine public health strategies,” wrote the authors of a report. “[They] exacerbate the spread of the pandemic or erode public trust and confidence in authority and government,”
Diverting resources to the technology “may undermine the pandemic response” the Ada Lovelace Institute warned.
Director Carly Kind suggest Britons may not adopt a contact tracing app with even the limited enthusiasm of Singaporeans. She notes that a supportive paper from Oxford’s Big Data Institute was co-authored by staff from the NHS contractor Faculty.AI – which is helping NHSX write the app – and questions its independence.
Faculty.AI was founded by Marc Warner, the chief executive, whose brother Ben was recently recruited to advise the government on data science. Ben Warner previously modelled the Vote Leave campaign and the Conservative’s 2019 General Election campaign.
Bonsall of the Oxford Big Data Institute rejects suggestions of a lack of independence and says that while “Faculty.AI are working with NHSX and we’re all talking to each other” the paper is “absolutely not pushing” the contact tracing project. The idea of using Bluetooth to fight a pandemic is not new and has support from serious computer scientists.
Almost a decade ago Cambridge University’s Professor Jon Crowcroft used the technology for the Fluphone experiment, which explored the feasibility of tracing an infectious virus. He says he was able to calibrate the system to an accuracy of around one metre, using less sophisticated tech than we have today.
Both Crowcroft and Bonsall argue there’s value if we define a “contact” not as a brief encounter but a prolonged meeting, of say 15 minutes.
“We’re not proposing a magic bullet. It’s about integrating it with all the other interventions, including manual contact tracing,” Bonsall adds.