Facebook’s patents show a commitment to collecting personal information

I personally assume that all news outlets are biased one way or another. And I do not believe there's a news outlet anywhere which is totally fair, unbiased and objective.

It seems reasonable to guess that the NYT's leadership believes FB is extremely influential in US politics and that it is imperative that FB avoid facilitating another 2016-style presidential campaign. In addition, most of the NYT's readership almost certainly believes that FB helped unfairly swing the election to Trump through illegal Russian funded ads and so forth, and it's good for business for a paper to cater to that readership.

I helped out a little with this piece. Aside from being happy to answer questions — I want to share a few bonus interesting Facebook patents that didn't quite make the final cut:

In "Systems and methods for directing messages based on social data" (https://patents.google.com/patent/US20180041464A1/en), Facebook talks about creating "user influence scores" as a metric of your reputation on the social network. It talks about how, "the user influence score can be decreased when the sender is reported to be associated ... with undesired content." Or can be decreased when your friends are "reported to be associated with undesired content."

In "Multi-Factor Location Verification" (https://patents.google.com/patent/US20170261590A1/en), Facebook ponders how to accurately track your location when you have your phone's GPS turned off. The phone can ping nearby RFID readers, NFC systems, see a nearby Bluetooth device, or the MAC address of your router, or the wifi SSID. But that's the easy stuff — what if you have your phone turned off entirely? The patent discusses tracking your location based on your calendar events, concert tickets, and dinner reservations: “A location determination method may include advanced knowledge of the user's intended or estimated location, such as a reservation for an event at a particular time (e.g. restaurant, theater, concert…)”.

In "Identifying users of client devices for tracking user interactions..." (https://patents.google.com/patent/US20170353564A1/en), Facebook discusses the problem of shared computers, where different family members may use the same web browser, and therefore the same tracking cookies. The patent discusses how to feed the web browsing and activity data into a "user identity matching module", which figures out which Facebook profile to attach the activity log to based on what you're doing online. Shopping for motorcycle parts? Log it for the husband. Dora the explorer? Probably the kid. Works just as well when you're logged out.

If you guys have any favorite mildly disturbing personal surveillance patents — please share! They're fun to collect.

And in related news — this piece running this weekend turned out to have interesting timing. This week the Supreme Court ruled, somewhat tangentially, on digital privacy, in Carpenter v. United States. The narrow majority opinion held that your cell phone tower location data should not be available to government search without a warrant. The five justices wrote about 24/7 location tracking: "Only the few without cell phones could escape this tireless and absolute surveillance."

And then some of the even more interesting argumentation came in the dissents — especially Gorsuch's.

He seems to want to develop an even stronger constitutional protection for personal data, throwing out the current "third party doctrine", and rooting it instead in a more expansive reading of the Fourth Amendment. He writes: "Can the government demand a copy of all your e-mails from Google or Microsoft without implicating your Fourth Amendment rights?" And "Just because you entrust your data — in some cases, your modern-day papers and effects — to a third party may not mean you lose any Fourth Amendment interest in its contents."

I know it's wishful thinking, but I'd love to see a day when "The right of the people to be secure in their persons, houses, papers, and effects, against unreasonable searches and seizures..." covered all of the personal data hoards that Facebook and Google have amassed.

We don't see people arguing we shouldn't patent physical machines because ultimately each operation is a simple physical movement, so the sum of everything is just a sequence of physical movements.

Yet we see people arguing we shouldn't patent software "machines" because ultimately each operation is a simple mathematical operation, so the sum of everything is just a sequence of mathematical operations.

The argument against the patentability of maths, AFAIK, has always been "maths is naturally inherent and merely discovered by humans". This, to me, is an argument against pure maths especially; but as we get into applied maths I think it's much less arguable that algorithms and data structures are naturally inherent (much like how physical constructions—machines—aren't naturally inherent from the laws of physics).

I still strongly think far too many patents get issued, but I'm less convinced than I used to be software patents are inherently bad, and that patents do need reform in various ways.