The "contact tracing" apps are actually "exposure notification" apps. Contact tracing is an incredibly labor-intensive, high-touch, face-to-face process with a long-proven track record in epidemics.
pluralistic.net/2020/05/02/rew…
1/
pluralistic.net/2020/05/02/rew…
1/
Automated exposure notification tools can be a useful adjunct to the work of contact tracing, though the apps that are being rolled out (including those built on the Google/Apple API) are untested.
2/
2/
Worse, they're operating in a data-vacuum for essential variables like "epidemiologically significant contact."
Making that number too large risks flooding users with false alarms that train them to ignore warnings, while going the other way will miss out real cases.
3/
Making that number too large risks flooding users with false alarms that train them to ignore warnings, while going the other way will miss out real cases.
3/
On top of that, there's the digital divide: the people who are least likely to have smartphones (poor people, old people) are more likely to get coronavirus (because of the runaway spread in precarious work environments, crowded shared homes and underesourced nursing homes).
4/
4/
They're ALSO more likely to experience severe and even deadly symptoms because of comorbidities like age-related ailments and poverty-related cardiovascular/nutritional/stress problems.
5/
5/
And finally, there's the looming risk that exposure notification apps will become permanent mass surveillance apps. This is exactly what happened after 9/11's "temporary emergency measures" were passed: 19 years later, most of them are still with us.
6/
6/
As it turns out, I'm not the only person who's ambivalent about exposure notification apps. Public health officials are also not that big on them.
wired.com/story/health-o…
7/
wired.com/story/health-o…
7/
Instead, officials in states like NY, CA, and MA, and cities like SF and Baltimore are rolling out tens of thousands of human contact-tracers.
Former/current federal public health officials, both R and D, want a national army of 180K tracers.
apps.npr.org/documents/docu…
8/
Former/current federal public health officials, both R and D, want a national army of 180K tracers.
apps.npr.org/documents/docu…
8/
They say addressing the pandemic will cost $12b for contact-tracers, $4.5b for quarantine housing in vacant hotels, and $30b in income support for voluntary self-isolators.
9/
9/
They've seen contact tracing work, and they believe in it, but they also understand that contact tracing's secret sauce is a person-to-person human trust-bond between the tracer and the subjects. It's hard, and hard-to-automate.
10/
10/
As @fvogelstein writes in @wired, academic epidemiologists are trained to ignore the traditional tech industry promises of an "easy" fix with a machine that replaces humans. And they're severely allergic to trying untested methods during emergencies.
11/
11/
Add to that the general bad odor that Silicon Valley has created for itself through toxic, monopolistic tactics, which is why politicians are less likely to go to the mattress to defend high-tech approaches.
12/
12/
There are signs that the tech industry is scaling back its ambitions, offering apps to manage the record-keeping and minutiae of manual tracing.
The alternative would be to go down the Chinese/SK/Singapore route of apps that jettison privacy protections.
13/
The alternative would be to go down the Chinese/SK/Singapore route of apps that jettison privacy protections.
13/
Not only are these apps impossible to square with US constitutional and temperamental constraints, but they'd also face the (insurmountable?) hurdle of being so mistrusted by large amounts of the public that they wouldn't be used widely enough to work.
14/
14/
What's more, the countries with "successful apps" ALSO had titanic numbers of human contact tracers laying down shoe-leather. Whether you think the US should or shouldn't do apps, there's no evidence that apps will work without legions of human contact-tracers.
eof/
eof/
