For the ML enthusiasts a nice semi-lay article about the difference between hype and reality by MIT Tech Review. Essentially a good review of the enormous effort that has gone in to COVID-19 targeted ML applications in the last 18 months, and that essentially none of them are fit for purpose. Its worth reading the MIT article if you have access, but they cite a Turing Institute report, a recent BMJ review, and a Nature review all highlighting overlapping issues - issues around data access, interoperability, lack of transparency or validation of models/data, amongst a range of other issues that mean that while a lot of time/money/effort went into ML modelling, not a lot of practical nor safe utility was published. I’ve attached the three referred articles in for your own reading. bmj.m1328.full.pdf (499 KB)
opengraphobject:[360669234110464 : https://www.technologyreview.com/2021/07/30/1030329/machine-learning-ai-failed-covid-hospital-diagnosis-pandemic/?truid=a110b7df5ef973db7c6b2799580c7516&utm_source=the_download&utm_medium=email&utm_campaign=the_download.unpaid.engagement&utm_term=Active%20Qualified&utm_content=08-01-2021&mc_cid=1769c307eb&mc_eid=a9afcedf98 : title=“Hundreds of AI tools have been built to catch covid. None of them helped.” : description=“Some have been used in hospitals, despite not being properly tested. But covid could help make AI better.”]