Read on Twitter
"We use TraceTogether to supplement contact tracing — not replace it." Great essay by @jasonbay, a pioneer of bluetooth smartphone app for contact tracing before Apple/Google h/t @kevinpurcell 1/ https://blog.gds-gov.tech/automated-contact-tracing-is-not-a-coronavirus-panacea-57fb3ce61d98
I've spoken with many who pooh-pooh good old-fashioned contact tracing by human workers interviewing patients. They say it's not scalable and will be replaced by apps. 2/
I don't agree, for multiple reasons. First, they often assume it's too expensive to hire human workers, so it's either automatic tracing or nothing. Actually cost of human labor would be minuscule compared to economic value of suppressing epidemic. Cf $2T stimulus package. 3/
Second, 7/10 Americans regard coronavirus as a "serious threat." I believe that some fraction of them would download an app, but the remaining 3/10 would be highly unlikely. 4/ https://www.reuters.com/article/us-health-coronavirus-usa-poll/forget-politics-majority-of-republicans-democrats-now-agree-coronavirus-is-serious-threat-reuters-ipsos-poll-idUSKBN21C02T
For reference, one in six Singapore residents has downloaded the TraceTogether app, an encouraging number but not close to 100%. 5/ https://www.straitstimes.com/singapore/about-one-million-people-have-downloaded-the-tracetogether-app-but-more-need-to-do-so-for
Google/Apple may have the marketing reach to do better, because of their app stores and control of smartphone OS. But I don't see how they will succeed with people who don't even regard coronavirus as a threat. 6/
Third, even if govts made app compulsory (which is obviously problematic), automatic tracing would not be able to extract all the information that a human can glean from an interview, nor could it make all the intelligent decisions of a human. 7/ https://blog.gds-gov.tech/automated-contact-tracing-is-not-a-coronavirus-panacea-57fb3ce61d98
Of course, over the long term, advances in the surveillance economy and artificial intelligence could make automatic tracing more capable. But remember that we need tracing to be effective starting in a few weeks when govts are trying to come out of lockdown. 8/
On that time scale, contact tracing tech will only augment human workers, not replace them. It should improve the speed and accuracy with which human tracers can do their jobs. 9/
So contact tracing tech is great, but don't let it distract you from the incredible urgency of hiring good old-fashioned humans. 10/
In the long run I'm bullish that COVID-19 will lead to a technological revolution in contact tracing. For example, cryptographic techniques like private set intersection could be used to protect privacy. 11/
Let me conclude by invoking Amara's Law: We tend to overestimate the effect of a technology in the short run and underestimate the effect in the long run. /end
I forgot to point out the technological progress with the greatest near-term impact in this area. @Accenture and @salesforce are bringing Customer Relation Management (CRM) systems to bear on contact tracing for the first time. /end+1 https://www.mass.gov/news/baker-polito-administration-announces-covid-19-community-tracing-collaborative-to-further
You know, the software that powers call centers. The tech is crashingly dull, but I've learned from hard experience that unsexy tech that you've never heard of, and no one wants to brag about, often has the greatest impact. /end+2
@marvinminsky once told me that someone, likely Sydney Brenner in the late 60s or early 70s, had tried to persuade him to apply AI to automate neural circuit reconstruction from electron micrographs, what we now call connectomics. /end+3
In reality, for decades the only impact of computers on connectomics was to assist manual tracing by humans via drawing software, which is considerably more boring than AI. /end+4
Just imagine the MIT AI Lab in the late 60s. You would have seen chess computers, theorem provers, and chat programs. You would have thought that AI was about to revolutionize the world. And you would have been wrong. Amara's Law. /end+5
You would also have seen some of the first text editors. Boring but word processing was one of the main drivers of the PC revolution in the 80s. /end+6
