Don’t Like What the Virus Models Say? Try Building Your Own (opens in new tab)

(bloomberg.com)

7 pointshannibalhorn6y ago2 comments

2 comments

2 comments · 1 top-level

hannibalhornOP6y ago· 1 in thread

I find the modelling discussion really interesting.

The Imperial College report that kinda kicked off mitigation efforts in the west was pretty bleak. I didn't question it myself, but just a few days later I saw a Bill Gates AMA where he said the model in that report used some parameters that were a bit too negative. [1]

Not long after, the lead author of the report announced he was getting help from Microsoft (and apparently even John Carmack?) to clean up the code involved. It's apparently several thousand lines of undocumented C written 13+ years ago. [2] The software engineer in me can't help but be skeptical as to how accurate it is in the current situation. Data science has changed quite a bit in the last decade.

1. https://www.reddit.com/r/Coronavirus/comments/fksnbf/im_bill... 2. https://twitter.com/neil_ferguson/status/1241835454707699713

SyneRyder6y ago

Carmack tweeted about the extent of his involvement:

"People are going to get the wrong idea about this from some of the headlines. I have been helping out for the last week as a “generic strong C programmer”, mostly doing static analysis, code quality, and resource optimization work. The code has been in development for a decade+."

[1] https://twitter.com/ID_AA_Carmack/status/1244302925855326209

j / k navigate · click thread line to collapse