It looks like California and, possibly, others are trying to put their focus into tracing and they want that to be viable before letting people go.
I’m not sure that I trust that the technology will be very effective. Hopefully, it will, obviously. But the issue I see is that say that I’m infected and infectious. Maybe I have symptoms, maybe I don’t. At best, I will have symptoms at some point but I’m still infectious for a couple of days before that point.
During the two days that I’m asymptomatic and infectious, I’m walking by people on the street, using the coffee machine at work, going to the bathroom at work, sneezing on things, etc. Some of those, the tracer app will pick up but some it won’t (to be fair, it looks like transmission of colds via objects is only about 2% - most transmission happens through sneezing and breathing on people).
If I actually get symptomatic, maybe I go to the doctor, maybe I don’t. But I might not ever become symptomatic - most likely I won’t. If I do go to a doctor or get tested, maybe a few more days have already passed. During all this time that I’m infectious, I’ve possibly interacted with dozens or hundreds of people. And the odds that I get symptomatic is fairly low (though, we’ll say 50% since I presume that there’s a correlation between infectiousness and likelihood of becoming symptomatic) and the odds that I am tested might only be 50% as well (quite possibly lower). So usually it won’t be me who get tested, it will be half the symptomatic people among those I came near, about a week later - after they’ve all had a chance to expose a few dozen or hundred people.
And well, now we know to quarantine a thousand or so people once those tests finally come back. But, among a thousand people, a pretty good quantity are going to be people who were at-risk. We just killed them because we told everyone that Google was going to use magic to make the disease go away and so they all decided that they can just live nice, normal lives again. Maybe they’re skipping shaking hands and doing a few small things, but mostly they’re still doing everything they usually would that helps to spread a disease.
And that all assumes that Google is watching everyone.
It looks like we have been tracking down illegal immigrants through their cellphones, so they have probably begun refraining from using smartphones.
Some people turn off GPS on their phone; some people don’t carry their phone around; sometimes service is bad; and so on.
A large vector of infection will be people traveling into the country from somewhere else. Their phone might not work here or it would be expensive to use, so they turn it off.
Overall, we’re going to have fairly significant gaps and that will compound the delay in detection by quite a lot and allow for some invisible Typhoid Maries to wander around. It will also make it impossible to identify people who were exposed in a not insignificant percentage of cases.
If your average person exposes 30 people and we only identify 80%, 6 won’t be discovered. We would expect 1 of those to become symptomatic and, possibly, more to become infectious but asymptomatic. Those 6 people might create another 60-90 exposures of whom, again, 20% will be undiscoverable. That’s enough error for a perpetual chain to not only continue but possibly even branch and expand.
Plausibly, we will get enough data that we can keep cutting off each outbreak at around 1-2k people and resetting the virus so that it has to find a new way in. But that really just slows things down. And, I don’t know that it slows it enough to protect at-risk groups from infection for an 18 month period. It’s good for the hospital system, but not for the people who will be killed in each of those outbreaks.
I think that tracing will make us feel good about ourselves. But then, when we go back and count the deaths from 2020, we’ll realize that we were deluding ourselves that our massive, heroic effort was doing the best that it could do.
I still believe that it is more practical, and likely to be more effective, to identify at-risk groups and focus on defending them. With tracing, you might not know that you failed until an at-risk person is dying in the hospital and you only discover that as ambulances are called out to the other hundred homes of at-risk person who were affected by the same outbreak.