I see people taking about “the”* COVID-19 contact tracking app again. I was part of a discussion of this on Friday.
* The app doesn’t exist, the legislation doesn’t exist, procedures don’t exist, and there’s no proper science.
* The app doesn’t exist, the legislation doesn’t exist, procedures don’t exist, and there’s no proper science.
I say “no science” because the claim that “If you use the app we can let you out” is an assertion, not a fact. Indeed, that the app approach delivers better quality data than people’s memories is an assertion not a fact. No one has has done this before. No one knows.
Read this: “Five ways a COVID-19 contact tracing app could make things worse” policyoptions.irpp.org/magazines/apri…
In brief, The App could prove useful as a supplement to memory. Or it could produce too many false positives. Or it could produce a false sense of security. Or other things. There doesn’t appear to be any modelling to tell us what’s likely here.
And in summary, it’s nothing more than Technology is Magick™ wishful thinking. We must do something. Here is something. We must do this. Wash your hands. Stay the fuck at home.
Oh. One more thing. All The App does is say you were within X metres of a known COVID-19 positive person for Y+ minutes. That’s literally it. It won’t know if you give your symptomless cousin a 30-second hug. It won’t know you slid your hand along a dirty handrail.
Infections don’t work like “Y-1 minutes safe, Y+1 minutes dangerous”. It’s all probabilities, and guessing what annoying behaviours people can stick to. A digital red light / green light is a bad way to communicate this.
There’s a reason the HIV/AIDS community spent decades stressing the term “safer sex” not “safe sex”. But you know, let’s not bother listening to people who’ve being doing messaging around infectious diseases for decades. Let’s save “the economy” with a magic app.
IMPORTANT CLARIFICATION on how The App actually works. It’s abstracted a LONG way from any social context. This is a thousand miles from how contact tracing actually works. It’s basically another Big Data Wet Dream.
When they say “measure distance” from signal strength they really mean “guesstimate distance based on possibly incorrect assumptions about surroundings that might absorb or reflect the signal”. It’s a known process but it’s also another source of probability/uncertainty.
FINAL NOTE FOR THE HARD OF THINKING: I’m not saying “The App is bad”. I’m saying there’s no real evidence to say whether it’s net benefit or not. It’s therefore wrong to say “Use the app and we’ll let you go out” without these risks being explained, which they’re not being.
Well here you go. Straight off the bat, what in my view is a real bad name. “Safe” as a binary. [Screams inside.]
Nor does it pick up a LOT of what human contact tracers do, like note that a bunch of positive people all bought something from the same coffee shop, or petrol from the same servo, over a 48-hour period. Because it quite properly doesn’t records location. Or it’d better not.
END OF THREAD. That’s my COVID-19 conversation quota done for today. No more replies from me today.
(Well, maybe if there’s solid, new information from actual experts.)
(Well, maybe if there’s solid, new information from actual experts.)
OK, just ONE more tweet on this thread. It’s out (at least the Android version). It’s called COVIDSafe. It has a website. I daresay there’ll by TV adverts from tonight. covidsafe.gov.au
I reckon @bengrubb and @joshgnosis will be good to follow on this. They’re on it now.
I reckon @bengrubb and @joshgnosis will be good to follow on this. They’re on it now.
@bengrubb @joshgnosis And ONE ONE more. The relevant legals. “Biosecurity (Human Biosecurity Emergency) (Human Coronavirus with Pandemic Potential) (Emergency Requirements—Public Health Contact Information) Determination 2020” legislation.gov.au/Details/F2020L… It’s a ministerial determination, so already law.
