[Paul-ish]

What is special about these feeds? The Facebook data was only publicly available data; ie stuff I could get by going to the right page in by browser. Was it the same with twitter? Why is it bad for a government to aggregate public data? Not just public data, but messages I assume the writers intend to be broadcast to the world. Were not talking about private emails or messages here.

I'm not trying to defend what is happening in a lot of areas with the police, I am just genuinely curious about the nature and potential for impact of these feeds.

[etherealmachine]

I'll bite.

> Why is it bad for a government to aggregate public data?

In the past, this has been problematic. There are several examples of government agencies using their authority to suppress political dissent. The Stasi in Germany, the KGB, Hoover's FBI. True, they had private data as well, but the problem wasn't the source of the data - it was that they used it for political suppression. It's clear that the agencies involved with Geofeedia would prefer the protesters just "go away" - they have no incentive to be supportive of their goals. Given the power imbalance and incentive misalignment, gathering information on them has an inherent smell to it, just from past experience.

[weberc2]

While I can appreciate the slippery-slope argument against mass surveillance and bulk data collection, comparing municipal police with infamous clandestine federal investigative agencies borders on conspiracy theory. While I don't doubt that the police dislike like the criticism, it seems more likely that their strongest incentive is to keep the peace (specifically preventing the destruction, theft, assault, and murder that characterize so many of these "protests").

[sedachv]

> While I can appreciate the slippery-slope argument against mass surveillance and bulk data collection, comparing municipal police with infamous clandestine federal investigative agencies borders on conspiracy theory.

If you look at the events of the last three years in the United States, municipal police have been much more aggressive in punitive actions against innocent people they perceive as inconvenient:

https://theintercept.com/2016/09/30/lawmaker-who-pushed-bill...

http://www.huffingtonpost.com/2014/08/03/ramsey-orta-arreste...

https://theintercept.com/2016/10/06/in-the-chicago-police-de...

This includes things like voter suppression at the request of the Republican party:

https://theintercept.com/2016/10/04/police-raid-indianas-lar...

[pavel_lishin]

> comparing municipal police with infamous clandestine federal investigative agencies borders on conspiracy theory

And yet: https://www.theguardian.com/us-news/2015/oct/19/homan-square...

Hoover's FBI is infamous because its practices were exposed. As we begin to look more closely at police departments around the country, we begin to see that there are many that end up using terrible tactics, and the only significant difference is that they're limited to their city and county borders.

[deftnerd]

One useful tactic that I use to give something "the smell test" is to reverse the targets. There have been cases where police acted improperly or illegally, so if this company was using cellphone data and aggregated social media data to let propestors know the identity and social media posts of all the police, would that be improper?

That feels wrong, so my "smell test" makes me also think that the other way around is wrong. Police should be acting on immediate actions of protestors, not potential actions based on some black box social graph algorithm that tells them who to arrest.

[weberc2]

> if this company was using cellphone data and aggregated social media data to let propestors know the identity and social media posts of all the police, would that be improper?

Yes, but only because the protesters aren't imbued with any special authority over the police (besides participating in a representative democracy) nor trained to use such a power responsibly. The impropriety has nothing to do with the corporation or the technology.

[sesqu]

So you argue that aggregating public data is still wrong, iff the publicizing party is an agent of an authority?

Would you then limit the impropriety to only those times when the agent is acting in official capacity upon that aggregating party, or do you think gathering social data about authorities should be blanket-banned until the aggregators take a class first?

In essence, I'm trying to figure out the implication of your use of the word "only". I suspect you intended it to qualify surveillance as a sanctionable, active operation, ignoring unordered or time-sensitive relationships and profiling.

[weberc2]

> So you argue that aggregating public data is still wrong, iff the publicizing party is an agent of an authority?

No, that's not my argument. My argument is that the test would always fail; the technological aspect is irrelevant.

My personal sentiment is that the police are probably using this technology appropriately, but I fear the slippery slope this precedent could set for federal agencies. I think that slippery slope is more dangerous than these protesters.

[whybroke]

No doubt the vast majority of police departments have good intentions.

But when the group protesting police malfeasance is then monitored by those same police in order to keep what those police consider to be the peace over them. And then they must rely on those same police not to target individual participants after the protest for retaliation, then that monitoring is obviously a very serious problem if the protesters be correct.

[M_Grey]

I really doubt that at this point. As a system, it's hard to ascribe good intentions to the behavior we've seen over the last few decades alone. The history of US policing is not a history of good intentions.

[weberc2]

Yes, the history curated by sensationalist media and revisionist historians is dark indeed.

[alasdair_]

One issue is that Facebook makes it relatively difficult to clearly know what data is completely public and what is not, especially for casual users.

In addition, a lot of the data is geotagged in ways that are non-obvious to non-technical users.