Apple/Google partnership immediate thoughts. From a user and efficacy perspective - Alice meets Bob. Thread/

Open sourced a model recently to show relationship between group “size” <> presence of asymptomatic. These apps will be...alert central (if you leave the house!). Folks working...will be around someone 70-95% of the time if group size > 300 https://docs.google.com/spreadsheets/d/1Vu0Yv-9mwMoCCMv2nfeG9heuknPYomC85rzzKjRjyE4/edit?usp=sharing

Why does that matter? Well... Keep in mind you’re only shown...nearby ‘someone’ - (anonymous)... of course Alice isn’t told Bob specifically is sick! Maybe she would want that info? But it doesn’t scale...publicly

Hmm possible failure points, rogue actor attacks? Well..just takes one clown to e.g. go in the subway see 10,000 people and then hit *infected* - alert 10K. These apps don’t show you “who” so it’s private. Questions: does Gov knock on your door when sick? Sick/clear toggle limit?

Btw any social graph - public or private - has these attack vectors. Apple and Google will design this well, but we can highlight these and more issues. Fun fact: Niantic’s Pokémon Go has to deal with massive threat of DoS and rogue actor attacks. Because...

Reality: if monetary incentive networks get attacked. Here weak $ incentive to use but $$ proxy is..strong health incentive. Can have bad actors (eek anon “bot” attack!) Not many will, but ex above, can happen w/ few. See also- multiple principal problem

https://en.m.wikipedia.org/wiki/Multiple_principal_problem