[Ancapistani]

Well, in my precinct I'd estimate there are ~20 people at the polls to vote at any given time. Given the timestamp of a ballot, there are maybe 50 people it could have possibly been.

That's more than enough information to correlate voting behavior after a couple of election cycles with a high degree of confidence.

Oh, and ballots aren't just for one race generally. By looking at what races that ballot voted in and a list of people present, there's a very good chance you'd be able to narrow it down to an individual in a single visit.

[Manuel_D]

The ballots are timestamped to a degree fine-grained enough to narrow down the set of potential voters to just 50 people? Which state is this? I'm not finding any search results indicating that ballots are timestamped. In fact, some of the results I've encountered specifically say that ballots aren't timestamped to ensure privacy. But happy to learn more of you can explain how ballots in your state are timestamped.

In my state, we have to sign our name on the envelope containing our ballot. If the government was corrupt and they wanted to identify how people voted, they could just look at the signature. No Flock required.

Of all the things to complain about Flock, the notion that it can somehow de-anonymize ballots is probably one of the most unusual I've heard.

[Ancapistani]

My point wasn't that Flock allows this, but that it allows an entire class of surveillance that was previously not available.

[Manuel_D]

What wasn't previously available? Automated license plate readers have been around since the 70s, and started being widely deployed in the 1990s as technology got cheaper: https://en.wikipedia.org/wiki/Automatic_number-plate_recogni...

It's not new tech.

[Ancapistani]

The closest thing to it was repo companies sharing data.

This is very much a new thing.

[etchalon]

At no point did I say it could de-anonymize ballots.

You claimed ballots provided the government with the information they needed to know who voted.

I pointed that is untrue. Ballots explicitly do not.

The fact you posted that tells you know have a Google level understanding of the law in the US, and the fact you posted an article about private citizens using public data as proof of the legality of government-operated mass surveillance data tells me you're a deeply unserious person who should probably read Robert's writing in the majority opinion in Carpenter.

[Manuel_D]

The 9th circuit upheld the use of automated license plate readers in US vs. Yang. The defense attempted to use Carpenter to argue against the legality of ALPR data, and failed: https://law.justia.com/cases/federal/appellate-courts/ca7/21...

I really appreciate the irony of you alleging a "Google level understanding" on my part, when your own argument was tried in a court of appeals and failed.

[etchalon]

If you're going to Google a rebuttal to sound smart, please read the opinion before you do.

The Ninth Circuit in US v Yang specifically did not rule on the applicability of Carpenter or whether ALPR's GPS database was sufficiently similar.

It ruled Yang lacked standing to sue on those grounds because you don't have any expectation of privacy in a rental car after you've turned it in.

It ... has absolutely nothing to do with anything.

I helpfully pointed out the actual case you should cite in a different comment.

Try Googling that one.

[Manuel_D]

But crucially, the police used the ALPR data without a warrant. Regardless of the rental car, the police did use ALPR data without a warrant and the court did allow that to be used in court.

It's still a court that came down in favor of warrantless use of ALPR data, even if the situation around the overdue rental car might limit it's application more broadly.