Friday, May 1, 2020

Contact tracing app: promoting widespread adoption

On February 28th, I made this suggestion to a friend who works at Google:

"Assuming it doesn't yet exist, how fast could Google develop/adapt an app to track face to face interactions that might be used to back and forward trace covid-19? Developing an accurate social network graph based on recall is hard, and having a real time data collection would allow the CDC to quickly trace those who have been exposed. What do you think?"

I elaborated on the idea about three weeks ago, when I wrote about an exit ramp from the shelter-in-place order.

"Technology could help here; an app if widely adopted that recorded all other apps in Bluetooth range proximity could help in that contact tracing effort, but like all apps with network effects, widespread adoption is needed to make it effective. Once such an app is developed, adoption might be accelerated if localities (towns, counties) required all restaurants to require the app be installed and active for all their patrons.

Another possibility is to require Fit-bit and similar health monitoring app companies to share data; this could be aggregated and annonymized giving a broad picture of small regions that may be heating up, triggering localized lock-downs; or it could be shared with personal data sent to the CDC (and the local authorities) to ensure that infected people are prevented from circulating freely in the community".

About two weeks after that, it was reported that Google and Apple were collaborating on an app to do exactly this - a contact tracing app.

One of the issues influencing its effectiveness is how widespread is its adoption. Google and Apple are solving the problem in a way that is likely to run into enormous backlash once people understand what they are doing; they are going to embed the app into the operating system and distribute it as the next update.

I predict the uproar will be enormous when people realize that tech firms are surreptitiously collecting data on their interactions and their health. Court challenges are inevitable. So we may need to fall back on voluntary adoption.

One way of making the app more attractive to users, and thus to promote its use without a mandate (or its inclusion by "default") is to enhance the value it delivers. As it is, the contact tracing data is useful to society as a whole but less useful to the individual participating in the data collection effort by using the app.  But what if the app could also tell you what the likelihood is that someone you are within 6 feet of has the virus? That would enable people to make their own real-time social distancing choices. Those concerned for their own health might find this a very useful feature.  Individualists would appreciate it, even if they were not concerned for their (or societies) health since it would allow for a relaxation of blanket restrictions on association. That could promote adoption if people saw use of the app as a reasonable trade off between monitoring and an extension of shelter-in-place.

No comments:

Post a Comment