[LINK] RFC: Does CovidSafe Actually Work?

[David Lochrin]

On 2020-04-29 19:18, John Mann wrote:

> I'm concerned about the effectiveness of solving the problem: finding everyone someone may have given the virus to.

I don't think it's intended to trace every contact.  It's effectiveness ultimately hangs on probabilities.

But even assuming the official target of a 40% uptake is reached and it's evenly distributed among the population (including those more likely to acquire CV19), the probability that a random close encounter will be detected is (0.4^2) or only 16%.  That would leave a large number of undetected contacts.

IMHO the main problem with this scheme is Government credibility re privacy.  Perhaps this could be overcome by entrusting the entire development, testing, and subsequent operation of the scheme to a trusted & independent organisation.  And the code should be open-source with an integrity check.

Of course the view from Link is not typical, and lots of well-meaning people will buy the marketing at face value.  Sigh...

David Lochrin