You get a really different picture on Twitter. And I'm not saying I know which one is right! https://twitter.com/GaryMarcus/status/1328048295109824512 https:/

You get a really different picture on Twitter. And I'm not saying I know which one is right! https://twitter.com/GaryMarcus/status/1328048295109824512 https://twitter.com/GaryMarcus/status/1328048295109824512 I follow Gary Marcus because I respect his views on AI, but I don't really appreciate the editorializing. One of the most horrifically inaccurate predictions in the history of public health. The obvious difference is that cases aren't deaths (and the amount of testing affects the number of cases, etc.). But if the plot is to be believed, then you could imagine that Sweden might NOT have excess deaths for 2020 (which is interesting in itself), but it could have many in the beginning of 2021. Or maybe almost all these cases will be mild and cause few deaths. There seems to be an extremely wide range of COVID outcomes.