Live data from Hacker News

Smartphone contact tracing has failed everywhere

eclecticlight.co

91–100 of 204 posts

Re: Smartphone contact tracing has failed everywhere

#91
post #11

Earlier quoted context omitted.

Over lives?

This is the same argument as justifying the NSA's mass surveillance with terrorism. Because lives are involved, it's propagated like a killer argument. Fact is, you don't know if even worse things aren't going to be enabled with this technology.

This technology is designed in the open in a privacy sensitive manner. Your phone tells far more to your cell phone provider, Apple/Google, all the apps providers you have, and credit card companies than this technology ever will.

Re: Smartphone contact tracing has failed everywhere

#92
> One of the potential advantages of a centralised model is that it would allow the use of machine learning to adapt thresholds to minimise error...

This is a cop out. Machine learning isn’t magic. If the proximity measurement doesn’t really work without ML then it’s probably not going to start working with ML. Besides, if you really wanted to do it in a decentralized app you could use federated learning, no?

Re: Smartphone contact tracing has failed everywhere

#93

I'm surprised this isn't mentioned in the article or any comments, and I'm wondering whether I'm missing something about the significance of this. But I thought the biggest problem was that Google and Apple didn't build a contract tracing app - they built a framework for local governments to build apps on top of. This leaves the work of marketing the app and driving adoption to each government (as well as building an…

There are two parts where local authorities need to be involved: troll-proofing the way people flag themselves as infected and translating exposure into actionable instructions/suggestions (e.g. different testing capacities require different testing strategies).

I agree that Google/Apple could have done a lot more, like an OS-branded App with extensive functional localization in cooperation with local authorities. But looking back at how pretty much every country was initially trying very hard to avoid working with the OS makers Google/Apple might not be to blame at all for merely providing a minimal API instead of a complete turnkey solution. Governments all over the world were only coerced into adopting the API by malware countermeasures already present in the operating systems, almost like a sudden vindication of Lessig's old "code is law" claim.

Re: Smartphone contact tracing has failed everywhere

#94

This is looking at it from a technological angle. However, what about the social/financial aspects? If I am told to self isolate for any duration of time, will the government provide 100% of my lost wages? Will they support businesses who could be put in real trouble from someone having time off? What about potential repeated cases of isolation? Most people (including myself) will do anything to avoid being on their…

All of this is true if you test positive with or without an app. The app helps you know you might be asymptomatic and spreading it to people that you know or don’t know, who can then die from it or otherwise be seriously injured.

Re: Smartphone contact tracing has failed everywhere

#95
post #75
post #67

Earlier quoted context omitted.

Indeed, if you start with the contact tracing platform, add a mass surveillance system to it, you end up with a mass surveillance system. Note that a surveillance system like the one you propose doesn’t need the contact tracing platform to work.

But it does make it a heck of a lot cheaper to build and scale. You need far fewer cameras and can spread cheap bluetooth beacons densely across the area. And you can put beacons in places you can't use cameras.

Wait, so how are you matching the beacons (cryptographic random values) back to individual users?The back end only gets that information if a user tests positive and agrees to submit the data.

Even then there is no reason to store that linkage so server compromises won't leak this information.

Re: Smartphone contact tracing has failed everywhere

#96
post #6

"Contact tracing" was the new "blockchain" for a while. It is a technology with great potential, but I don't expect a truely effective, well-designed solution to be deployed at a large enough scale to be useful for years to come. I think it will eventually mature enough to be useful for future outbreaks - even for more benign disease. For now, though, I want to see the technology spend a little more time in the oven.

Do you actually have specific criticisms or is this just a vague feeling? To say it’s not well thought out is a baseless accusation.

Re: Smartphone contact tracing has failed everywhere

#97

I'm surprised this isn't mentioned in the article or any comments, and I'm wondering whether I'm missing something about the significance of this. But I thought the biggest problem was that Google and Apple didn't build a contract tracing app - they built a framework for local governments to build apps on top of. This leaves the work of marketing the app and driving adoption to each government (as well as building an…

I think the individual apps on top of the same framework are not identical in the way the data is collected and stored. AFAIK in some countries they have a centralized database, whereas e.g. in Germany all contact IDs are kept locally and only compared to a published (but anonymized) list of infected IDs.

No, that's not the difference. The contacts are stored inside the system's framework.

The difference is in how people report themselves as infected. That has to match the country's process for verification of the result.

Re: Smartphone contact tracing has failed everywhere

#98

I'm surprised this isn't mentioned in the article or any comments, and I'm wondering whether I'm missing something about the significance of this. But I thought the biggest problem was that Google and Apple didn't build a contract tracing app - they built a framework for local governments to build apps on top of. This leaves the work of marketing the app and driving adoption to each government (as well as building an…

Ask me if I will volunteer to give apple or google any more information than they already collect and my answer will always be no. I don't care what the reasoning. Germany did it right.

Since the contract tracing api doesn’t give information to Apple or Google, it’s hard to see what relevance this comment has.

Re: Smartphone contact tracing has failed everywhere

#99

I'm surprised this isn't mentioned in the article or any comments, and I'm wondering whether I'm missing something about the significance of this. But I thought the biggest problem was that Google and Apple didn't build a contract tracing app - they built a framework for local governments to build apps on top of. This leaves the work of marketing the app and driving adoption to each government (as well as building an…

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?

Re: Smartphone contact tracing has failed everywhere

#100
I live in Ireland and our contact tracing app (using Google+Apple tech) seems to be doing well. There was about 1 million install 1-2 days after it launched, currently standing at 1.4 million installs (population is 4.9 million), along with some positive contacts reported:

https://www.bbc.com/news/technology-53525712

It was also donated to open source.

https://www.nearform.com/blog/ireland-donates-contact-tracin...

While that is useful I think each country will face its own challenges with adoption.

Post reply on HN