Smartphone contact tracing has failed everywhere • Hacker News Board

Re: Smartphone contact tracing has failed everywhere

#101

Post by johannes1234321 » Sat, Jul 25, 2020, 5:01 PM UTC

Example of the success of the German app: https://www.lra-bgl.de/t/das-landratsamt/aktuelles/details/n... (sorry German)

In the region of Berchtesgaden a asymptomatic person for a warning from their app and was confirmed positive. Due to the app that person didn't spread further.

Now we can argue how many of such cases we need for calling it a true success and then we can discuss if a single case makes the news if that's a bad sign ... on the later it's important to mind that certainly not every case makes that news and it's from a region with low virus load to begin with.

Re: Smartphone contact tracing has failed everywhere

#102

Post by zepto » Sat, Jul 25, 2020, 5:04 PM UTC

makomk wrote: post #60

DangitBobby wrote: post #45

Earlier quoted context omitted.

Specifically, it seems that adoption rates are so low as to render the application nearly useless.

Yeah, the overall conclusion is that smartphone contact tracing probably isn't working, but we can't know for sure because the megacorps who designed it did so in such a way that makes it almost impossible to measure if it's actually working. The only app we know has definitely been ineffective is Australia's centralized solution, in large part because they do actually have the ability to collect statistics on its ef…

It’s trivial to tell whether it’s working with some ordinary surveys and research.

No “megacorps” have done anything to make it impossible.

Re: Smartphone contact tracing has failed everywhere

#103

Post by goblin89 » Sat, Jul 25, 2020, 5:07 PM UTC

sassypotato wrote: post #55

neuronic wrote: post #43

Earlier quoted context omitted.

Asia has solved a lot of the issues surrounding the pandemic months ago. While American and European economies will be crushed by Corona (it's I bet it's not even half time) due to conspiracy theories, ignorance and egoists and adult babies crying about wearing a mask, Asia is getting back to ramp up their economies. They are far more capable of shutting new clusters down AND CONTROL IT. When we send people into quar…

Yes such a shame we can't focus on our economy, just because of outdated things like liberty and human rights.

The Economist’s Democracy Index as of 2019 rates favorably enough most East Asian democracies that have been handling COVID-19 well so far; some of them score higher than the US or half of European countries.

It doesn’t seem like “capable of containing a highly infectious disease” and “oppressive” go hand in hand.

Re: Smartphone contact tracing has failed everywhere

#104

Post by usrusr » Sat, Jul 25, 2020, 5:07 PM UTC

I disagree with the failure claims except for lack of adoption.

How can unreliable distance measurements be an issue? Actual viral spread over distance is a much, much bigger uncertainty. Even improving distance measurements to the nanometer wouldn't make much of a difference considering transmission prediction. Sometimes the virus bridges tens of meters, other times it fails to make landfall in a prolonged skin on skin situation.

Measuring impact is a simple polling research problem: if you want to know this kind of thing (and I think you should) you need to accompany your PCR tests with a tiny little questionnaire collecting data about why the test was done. Failure to collect that data is entirely unrelated to the app.

Re: Smartphone contact tracing has failed everywhere

#105

Post by ridewinter » Sat, Jul 25, 2020, 5:12 PM UTC

zepto wrote: post #99

ridewinter wrote: post #84

Earlier quoted context omitted.

The whole reason for Apple-Google getting involved in the first place was a technical issue that prevents Android from seeing iOS. Instead of solving that issue to allow a variety of solutions to be developed, they forced the entire world to create the apps in a certain way. It's not that their privacy-centric approach was wrong (although this article details its major shortcomings), but the suppression of trial & er…

If there were no gatekeepers on this, how much contact tracing fraud and malware would there be?

These would be good problems to have because they'd mean the apps were successful in the first place! Problems are solvable if we have the chance to solve them.

Re: Smartphone contact tracing has failed everywhere

#106

Post by tzs » Sat, Jul 25, 2020, 5:22 PM UTC

gruez wrote: post #62

jmkerr wrote: post #42

Earlier quoted context omitted.

Random Bluetooth identifiers rotate every 10-20 minutes, to help prevent tracking [1]. (Of course it's always possible to track phones with their radios on for some time and distance.) [1] https://covid19-static.cdn-apple.com/applications/covid19/cu...

That's irrelevant because once you're infected, you give up your daily keys[1], which allow your random ids to be correlated together. [1] technically this is optional, but not doing so also means you're not reporting your infected status to others, which kills the point of contact tracing.

The medical facility that diagnoses you does not get your keys, if I'm understanding the system correctly. They give you a key that you give to your phone, and your phone then uploads the daily keys. Unless your phone includes identifying information with that upload, the daily keys are still anonymous at that point.

The hypothetical malicious actor would need to have (1) listened at some location where you were and captured the key you were broadcasting at the time, (2) identified you by some other means, such as face recognition, while you were at that location, and (3) infiltrated the server where your phone uploaded its daily keys so they could find out what other keys came from the same device as the one they captured and correlated with you already.

If they do all that then they could tell which other locations they were capturing key broadcasts at that you visited.

If they have not gained access to the data at the upload server, then all they get from matching your key to your real identity at one location is what other key listening location you visited before the key rotated.

They could put the face recognition systems (or whatever they are using) at their other key listening stations...but then they don't need the keys.

This seems like a lot of effort for little gain on the part of the malicious actor.

Re: Smartphone contact tracing has failed everywhere

#107

Post by Barrin92 » Sat, Jul 25, 2020, 5:24 PM UTC

anoncake wrote: post #79

Barrin92 wrote: post #51

Earlier quoted context omitted.

The false choice here is actually that you can sacrifice lives and still expect to have freedom. The alternative to not giving up privacy for contact tracing or similar measures is to be locked into your home medieval style while the plague runs wild in the streets, which is a significantly stronger limitation of freedoms. The choice isn't between liberty or safety, just as it isn't between economic health and lockdo…

Locking yourself up for fear of a <1% chance of dying is very much a choice.

The chance of ending up in a hospital is much higher (about 15 times higher), and so are the chances for complications in general like loss of smell or neurological or heart/lung damage. And if you have pre-existing conditions (like a sizeable amount of people) those risks only go up.

As can be seen in slumping economic stats many people are avoiding that risk which makes complete sense. It's absurd to frame this as a 'choice', it's a gigantic societal and civilisational failure to not be able to stop this disease.

In particular in the US though framing collective failure of preventable harm as cost for freedom has become the norm. You can pick anything that doesn't work or that could have been prevented and within five seconds someone will tell you it's the price for freedom. I think the real explanation is much more sinister. As a society people have given up, so now every harm or burden is rationalised in this way.

Re: Smartphone contact tracing has failed everywhere

#108

Post by makomk » Sat, Jul 25, 2020, 5:31 PM UTC

thetanil wrote: post #81

nxpnsv wrote: post #78

Earlier quoted context omitted.

So they’d have to just force it then?

Or do it like Germany, no centralized info store.

The German approach doesn't give your information to a German centralized info store, but apparently it does require agreeing to feed your location data to Google's own databases on Android devices: https://www.nytimes.com/2020/07/20/technology/google-covid-t...

Re: Smartphone contact tracing has failed everywhere

#109

Post by denysvitali » Sat, Jul 25, 2020, 5:31 PM UTC

The thing that still blows my mind about the whole contact tracing solutions in Europe is that the servers do not federate with themselves, and that you can only keep a single country specific app active at the time.

This leads to the situation where people that usually cross the country's borders will only use a single app which may not track all of the contacts. So far, the current solution is IMHO useless - but I keep on using it.

In my case I just keep one app active at all time (the one from my home country), but it would be more effective if with a single app I was able to track the contacts with people from other countries too.

Stupid simple example: I live in Switzerland (SwissCovid), and ideally when I cross the italian border I'll need to activate the italian contact tracing app (Immuni), but contacts with people that use a different app won't be counted even if they are infected - which totally misses the point of having a contact tracing in the first place :(

Re: Smartphone contact tracing has failed everywhere

#110

Post by bad_user » Sat, Jul 25, 2020, 5:33 PM UTC

Barrin92 wrote: post #51

bad_user wrote: post #17

Earlier quoted context omitted.

It's a false choice. But even if it were true, yes, over lives. People already died fighting for my liberal rights, I'm not willing to trade any of that for some temporary safety. I'm from a Warsaw pact country. Seeing westerners so willing to give up their rights these days is so deeply disappointing.

The false choice here is actually that you can sacrifice lives and still expect to have freedom. The alternative to not giving up privacy for contact tracing or similar measures is to be locked into your home medieval style while the plague runs wild in the streets, which is a significantly stronger limitation of freedoms. The choice isn't between liberty or safety, just as it isn't between economic health and lockdo…

While I understood lockdown measures, know that this is your privilege talking. You can stay in lockdown because you're middle class living in a wealthy country. In countries like India lockdown can cost far more than what Covid-19 can do.

I'm not arguing against public health measures, when the cost to me is low or when accompanied by evidence.

So where's the evidence that tracing apps work? That South Korea does it and South Korea manages to keep Covid-19 under control, that's not evidence. Plenty of confounders like the massive testing they did, their discipline in wearing masks, etc.

And note that I'm not a denier. I do think 1% mortality is unacceptable, plus it appears that's not the only problem. High hospitalization rates, permanent heart and lung damage, etc.

Still not willing to trade my privacy away, sorry.

You're free to do that but I'm willing to fight for mine.