There are SO MANY teams right now working on contact tracing apps that use bluetooth and GPS logging so they can alert you when you've been in contact with someone that tests COVID-positive.

My hot take: they won't work. For my first-ever tweetstorm, I will tell you why. [1/n]

The effectiveness of contact tracing is a quadratic function of adoption. To trace contact between two people, they BOTH need to have a specific contact tracing app installed (and bluetooth/gps needs to be on, and the phones can't be buried too deep in a bag, and and and..) [2/n]

The best team that has implemented contact tracing so far is Singapore. Open-source, privacy preserving, and marketed by the government. It saw record-breaking downloads for a government app. Today, 1 in 6 Singaporeans have it installed. Pretty good, right? [3/n]

Not good enough. The chances of two people in contact having the app installed is (1/6) * (1/6) = less than 3%. And that's an app sponsored, sanctioned, and pushed by the government. I think an app from MIT/Stanford/YC/whatever is going to see even lower adoption. [4/n]

Worse: on iOS, it only works if the phone is unlocked and the app is in the foreground. WHAT! Yes: the government tells you to turn on the app and flip your phone upside down. Even 3% is highly optimistic. [5/n]

Contact tracing is nerd-bait. So many scrappy teams are working towards this promise to solve the virus with an app. Only an official effort, led by Apple+Google or maybe FB and then forced upon users, can reach the critical mass needed to make contact tracing viable. [6/n]

Ok so, I bet one of those three companies will release a contact tracing app by the end of May, and the internet will probably be angry and scared of it, but maybe we give it a chance so we can finally go outside again? [7/fin]