[00:16.620 --> 00:17.660]  People have filtered in.
[00:18.620 --> 00:26.480]  My name is Eleanor Saitta and I'm going to talk a bit about surveillance today and the way that surveillance affects our lives.
[00:28.860 --> 00:33.940]  There are a lot of different kinds of surveillance that we interact with in a modern city.
[00:34.720 --> 00:37.040]  We've been watching people for a very long time.
[00:37.300 --> 00:46.640]  And every time we invent a new form of media or a new technology, we also eventually invent a new kind of surveillance that goes along with that technology.
[00:47.020 --> 00:53.080]  The list you see here of surveillance technologies is almost necessarily incomplete.
[01:05.010 --> 01:16.010]  So, to make this a bit more real, let's take a minute to look at how many ways an average city dweller interacts with surveillance and monitoring systems in a sort of prototypical city.
[01:18.530 --> 01:24.090]  So, you wake up and you turn on the lights and a smart electric meter notes the change in power use.
[01:24.470 --> 01:26.470]  While you make your coffee, you check your webmail.
[01:26.710 --> 01:29.170]  Your ISP screens and stores all of the web traffic.
[01:29.470 --> 01:34.310]  You reply to an email and the intelligence agency stores it and keyword matches the contents.
[01:34.590 --> 01:38.370]  Your mail provider analyzes the information and sells it to an ad company.
[01:38.910 --> 01:42.630]  You turn on your phone and it registers your location with the E911 database.
[01:43.170 --> 01:44.790]  The phone company also records it.
[01:44.790 --> 01:50.090]  The intelligence agency cross-references the two new text messages that come in with the email you just sent.
[01:50.430 --> 01:53.590]  You take a shower and the gas and water meters note the usage.
[01:53.950 --> 01:56.370]  You get dressed and get on the subway to go to work.
[01:56.670 --> 02:04.130]  Along the way, you appear on cameras run by your landlord, the transit agency, several nearby merchants, the police, and the traffic agency.
[02:04.850 --> 02:08.650]  The time you swipe your card to enter the subway is stored, along with your identity.
[02:10.890 --> 02:13.390]  While you're on the platform, you get a call from your boss.
[02:13.730 --> 02:18.390]  The numbers, time, and duration are logged for billing use and possible law enforcement use.
[02:18.710 --> 02:28.510]  The phone company doesn't capture the voice data, but the company that you work for may capture it on his end, and the microphone attached to the camera that you're standing under might also capture it.
[02:29.150 --> 02:32.730]  The train arrives, and you walk by an air quality monitoring box as you get on.
[02:33.810 --> 02:40.490]  As the train leaves the station, the transit authority tracks the train's movement, and onboard cameras record you.
[02:41.350 --> 02:45.890]  When you walk out, the destination of your journey might be recorded in your MetroCard with some systems.
[02:46.490 --> 02:50.630]  The turnstile also increments the number of people leaving the station in this time period.
[02:51.990 --> 02:57.930]  On the way into the office, in addition to passing through the vision of a couple dozen more cameras, you stop to buy a croissant.
[02:58.170 --> 03:05.270]  You pay by credit card, and the transaction is recorded by the merchant, the credit card processor, the credit card company, and the issuing bank.
[03:05.450 --> 03:10.450]  The last three of them all run it through fraud detection systems and sell data to marketers.
[03:11.150 --> 03:14.950]  Your mobile registers your changing location as you walk past a weather station.
[03:14.950 --> 03:21.010]  The radio signal from your phone is evaluated by an RF threat detection system run by the National Security Agency.
[03:22.670 --> 03:32.370]  A cab with its own video system and GPS tracker drives by, and plate and face detection systems recognize the car and the driver separately.
[03:32.870 --> 03:39.030]  It backfires, and its onboard data collection system for its emissions capture a whole bunch of data from that.
[03:39.310 --> 03:44.350]  A nearby gunshot-detecting microphone evaluates that noise and throws it out as a false positive.
[03:45.110 --> 03:53.010]  The car itself is also logging engine data and speed data, plus acceleration, which the police and insurance companies can use in the event of an accident.
[03:53.470 --> 04:02.190]  It drives over a temporary traffic load sensor and past a red light camera, and less obviously, past a chemical sniffer looking for explosive residue.
[04:03.090 --> 04:10.390]  You walk into your office, through an alarmed door, and past a window with a glass-break detector, and also past a heat rise and smoke detector.
[04:10.870 --> 04:16.950]  You show up on another camera as you swipe your card to get in, which is logged, as is the time you log into your computer.
[04:16.950 --> 04:20.390]  You load up a web page, and that's checked by a corporate web filter.
[04:20.890 --> 04:22.750]  You've been awake less than two hours.
[04:23.030 --> 04:28.870]  Half a dozen government agencies have logged information about you, plus at least three times as many private organizations.
[04:31.950 --> 04:45.410]  So, now that we have some idea of the scope of the issue, and just how pervasive it is, let's take a step back and look at some of the structural issues around surveillance to better understand its position within this larger socioeconomic frame.
[04:46.150 --> 04:55.570]  One of the distinct complications of looking at surveillance work is the same property of digital information that makes it fundamentally a more useful medium for its intended purpose.
[04:55.870 --> 04:58.950]  It can be really easily retransmitted and aggregated.
[04:59.670 --> 05:08.310]  The physical installation of sensors is obvious and visible, but once those sensors exist, repurposing that data can be done much, much less visibly.
[05:08.310 --> 05:19.210]  For instance, when a municipality decides to install a camera system, they may first present it as a traffic camera, theoretically only in place to watch for accidents and to review traffic levels.
[05:19.670 --> 05:25.050]  However, the police very quickly shift those same traffic cameras over to more general use.
[05:25.150 --> 05:29.070]  This has been true since CCTV was installed anywhere.
[05:29.070 --> 05:37.490]  The very first system in Munich in 1958 was originally a traffic camera system that was used for general police work within a couple of years.
[05:37.970 --> 05:41.050]  The pattern repeats itself for many different technologies.
[05:41.390 --> 05:44.110]  Mobile phone location tracking is a great example here.
[05:44.690 --> 05:55.830]  Originally in the U.S., the E911 location reporting feature was only designed to let emergency operators know where people were calling from, because this was an increasing problem that they were having.
[05:56.110 --> 05:59.850]  People would call up with cell phones and they wouldn't know where to send the ambulance to.
[06:01.510 --> 06:05.490]  However, police and intelligence agencies rapidly, rapidly adopted it.
[06:05.670 --> 06:24.810]  Last year, one out of the four major U.S. carriers reported more than 8 million cell phone location data requests just for GPS data, not for tower based location data that's much more common, not for 911 calls, not for intelligence agency operations, not for civil lawsuits,
[06:24.890 --> 06:26.330]  just from the police.
[06:26.730 --> 06:28.410]  And that's just one carrier.
[06:29.610 --> 06:38.790]  So, the largest forms of secondary use, even larger than that kind of thing, involve sharing and buying and selling data.
[06:39.150 --> 06:42.870]  This is a pretty fundamental feature of modern society.
[06:43.510 --> 06:51.070]  The very concept of a credit rating is based on companies sharing financial information with each other through mediating rating agencies.
[06:51.330 --> 06:54.050]  This happens both for consumers and for large enterprises.
[06:54.530 --> 07:01.690]  The same thing is now happening with other less clear cut information, like the results of data profiling performed by marketing firms.
[07:02.210 --> 07:14.050]  An amazing amount of information is included in these profiles, down to stuff like mood data reported by voice stress analysis systems that are built into interactive voice response systems.
[07:14.650 --> 07:24.650]  So, when you call up some kind of thing and it says, you know, oh, say operator for operator, it tries to figure out what kind of mood you're in and that affects how your call gets routed.
[07:25.010 --> 07:34.310]  You know, sometimes if there's a menu system that isn't recognizing you and you just start screaming at it, it'll transfer you to an operator even though it didn't recognize anything.
[07:34.310 --> 07:39.210]  But it also takes a note of who you are once it figures that out and ties into account data.
[07:39.670 --> 07:43.030]  And a lot of times that stuff sticks around on those records and then gets aggregated.
[07:46.090 --> 07:48.250]  So, you know, and this can include stuff.
[07:48.410 --> 07:55.470]  So, you know, you have a marketing profile that then says, oh, the last time you called, you know, the last five times you called your credit card company, you've been really upset.
[07:56.030 --> 07:58.690]  Like much more so than their average users are.
[07:59.750 --> 08:04.670]  In much of the world, these kinds of data and clearing houses are really largely unregulated.
[08:05.950 --> 08:07.850]  Similar things exist within the state.
[08:08.170 --> 08:17.150]  In the past 10 years especially, governments have been building large interagency databases and they collect all sorts of information, not just about convicted criminals.
[08:17.890 --> 08:24.630]  National intelligence agencies have also been buying, have taken to buying the data from civilian clearing houses.
[08:24.630 --> 08:34.290]  They can't economically collect all that data on their own, but that also means that they can get around the rules on who they can watch and what kinds of data they can collect.
[08:35.650 --> 08:46.110]  But that means that they get data that was never vetted for intelligence use, was never intended for that kind of, you know, for the kinds of roles it gets pushed into, like those voice stress systems.
[08:48.550 --> 08:58.070]  These kinds of data sales and really the entire secondary use system are driven by, among a lot of other things, but primarily driven by the expense of gathering this data.
[08:58.790 --> 09:06.670]  Once the city has spent millions and millions of dollars on a CCTV system, there's immense pressure on the city to make the most of that investment.
[09:06.950 --> 09:09.990]  This leads to second uses of that data being set up.
[09:10.250 --> 09:17.970]  These secondary uses may not have much, if any, direct benefit to the general population, but they show that the city is making the most of its investment.
[09:18.270 --> 09:29.830]  You know, a lot of the time these things are done either for internal city political reasons, that, oh, it would be really convenient if we had this, or they're just done to literally, you know, show the voters, oh, well, we spent all this money, but look,
[09:29.930 --> 09:31.510]  it's being used for all of these things.
[09:31.950 --> 09:34.630]  Whether or not anyone actually wants that, it still gets the message through.
[09:35.470 --> 09:45.210]  A similar pressure exists in the commercial world, where companies that have accumulated data for legitimate primary business purposes look for ways to defray the cost of all that data accumulation.
[09:45.470 --> 09:48.670]  This is one of the big things that happened with credit card companies.
[09:50.290 --> 09:52.890]  You know, that, oh, we have this huge database.
[09:53.050 --> 09:55.770]  Well, how can we make money on that huge database?
[09:56.010 --> 09:57.130]  And then they start selling it.
[09:57.790 --> 10:12.870]  Similar pressures are resulting in a lot of public-private cooperations, where you get networks of public and private cameras going together, that kind of thing, or the places where intelligence agencies are buying databases from the marketers.
[10:14.310 --> 10:22.210]  The same pressures that cause this second use end up being behind a lot of the cases where the government uses surveillance abusively.
[10:23.430 --> 10:33.450]  If you have the opportunity to do this sort of thing, you know, if you have the opportunity to abuse a surveillance system, that will happen simply because, oh, it's easy.
[10:33.570 --> 10:34.290]  Why not do it?
[10:34.290 --> 10:44.870]  It's really pretty rare that governments of, you know, so-called free countries, free nations, specifically go out of their way to build systems to track, for instance, political dissidents.
[10:45.070 --> 10:49.730]  That's really kind of hard to justify if you actually build the system with that in mind.
[10:49.990 --> 11:08.490]  However, it's a regular, regular occurrence that systems that are built for some legal purpose, whether they're wiretap systems with insufficient safeguards, databases that are only intended to track known criminals, et cetera, are used to track and to monitor and to harass dissidents and both political and even personal enemies.
[11:09.670 --> 11:13.110]  The more global a system's reached, the more damage it can do.
[11:13.370 --> 11:21.290]  The more easily that data can flow around through a system, the harder it is to ensure that there are appropriate controls on all levels.
[11:21.290 --> 11:29.990]  The progress of the technology of surveillance is directly at odds with the prevention of abuse, even assuming that that surveillance is good in the first place.
[11:33.630 --> 11:39.850]  So, one of the other issues is the equality of surveillance versus aggregation of data.
[11:40.370 --> 11:53.870]  There's a large class of initiatives that either do invade people's privacy or have the potential to do so that are not based on gathering new data, but just based on making data which already exists more available.
[11:54.710 --> 12:00.430]  Even if it's, you know, available in some form, that doesn't mean that it's easy to get at.
[12:00.810 --> 12:08.570]  The argument is that there's no actual change in how much privacy someone has when this happens, but that's pretty obviously false.
[12:09.330 --> 12:11.930]  Privacy isn't an absolute binary function.
[12:11.930 --> 12:20.710]  It's a gradation across types of information, across different amounts of data that are available, and most specifically across ease of access.
[12:21.030 --> 12:26.590]  Changing any of these has an immediate and concrete effect on individual privacy.
[12:27.410 --> 12:37.870]  This alone, assuming that we say that violating someone's privacy is an evil action, that there is a direct social harm there, kind of puts to lie Google's claims to not be evil.
[12:37.870 --> 12:52.130]  I mean, no matter how much good they've done in other cases, they have a very, very shaky record on personal privacy, and they've done more than any other single corporation to make all kinds of data, regardless of, you know, what it's used for, what it is,
[12:52.230 --> 12:54.310]  anything like that, easily available.
[12:56.650 --> 13:01.090]  So when you get things like aggregation and data profiling, stuff gets even more complicated.
[13:01.090 --> 13:14.830]  When companies or agencies make an effort to collect large swaths of information from both public and private sources and create a profile out of it, they get a lot more information out of that after it's been processed than just what they captured.
[13:15.130 --> 13:24.870]  They can do things like say, well, in 95% of the cases where we saw this pattern of purchases, it meant that someone had just lost their job.
[13:24.870 --> 13:32.030]  Now, chances are maybe you are in that 5%, but they can just simply say, well, yeah, this is probably what that means.
[13:32.490 --> 13:40.430]  Small, fairly trace amounts of data can be statistically telling to a really surprisingly... to a really surprising degree.
[13:41.050 --> 13:43.330]  Economies of scale help this process a lot.
[13:43.550 --> 13:55.790]  Both you need the economy of scale to build out the system and get the database that you're doing your analysis from, but also to build a system like that to investigate one person would be really horrendously expensive.
[13:56.050 --> 14:03.390]  There's no way you can justify spending a hundred million dollars, even if you're the NSA, to investigate a single person.
[14:03.650 --> 14:09.850]  However, a hundred million dollars to investigate a hundred million people, okay, sure, why not?
[14:11.370 --> 14:18.530]  This difference in economies of scale is one of the reasons why the future that we all go along with having no privacy is pretty fundamentally unworkable.
[14:19.050 --> 14:22.050]  Even assuming there isn't really an intrinsic need for privacy.
[14:22.410 --> 14:26.690]  It doesn't result in the equality that it's being sold as resulting in.
[14:28.090 --> 14:31.510]  Differential resources create a power dynamic in who can know what.
[14:31.770 --> 14:37.630]  If the results of all of those data profiling systems were public, then there might be an argument, but they're not.
[14:38.150 --> 14:45.410]  There's no reason and there's no way to compel some marketing firm to release all of the data profiles that they just spent all of this money on.
[14:45.410 --> 14:50.410]  And unless that information is public, you don't actually have that kind of equality of privacy.
[14:51.630 --> 14:56.810]  Beyond that, organizational privacy and personal privacy are completely different things.
[14:57.110 --> 15:03.730]  You can hurt a corporation monetarily when you leak data about, you know, what's going on internally, that kind of thing.
[15:03.850 --> 15:06.070]  But that doesn't cause psychological harm.
[15:06.310 --> 15:09.070]  You cannot hurt a corporation's feelings.
[15:11.090 --> 15:14.450]  The legal system is also really good at taking care of corporations.
[15:14.770 --> 15:19.750]  The kinds of privacy that corporations care about have specific legal protections.
[15:20.010 --> 15:21.970]  This is why we have trade secret laws.
[15:22.150 --> 15:24.310]  This is why we have copyright laws.
[15:24.450 --> 15:26.790]  This is why we have patent and trademark laws.
[15:26.910 --> 15:36.830]  These are all weapons that they can use to stop information from going public or, you know, modulate how disclosure happens and how information is used.
[15:37.810 --> 15:43.330]  That same lobby, you know, they spend a lot of money trying to make sure that those laws are protected.
[15:44.030 --> 15:45.810]  Personal privacy doesn't have that lobby.
[15:45.970 --> 15:47.630]  It doesn't have that money getting spent for it.
[15:50.710 --> 16:00.010]  So, now that we have a better handle on some of the things that make surveillance complicated, let's look at some more specific examples of how surveillance can affect people.
[16:00.910 --> 16:05.150]  Most people tend to think of surveillance as a bad thing and that's how we've mostly looked at it.
[16:05.270 --> 16:11.390]  But there are ways it can actually be really good depending on who you are and what your relation to the surveillance system is.
[16:12.190 --> 16:19.430]  Because the effects of any individual real world system are so complex, we're going to look at specific facets of different systems.
[16:19.850 --> 16:26.070]  Some of the real world implementations of these systems have different effects, have secondary effects and that kind of thing.
[16:26.070 --> 16:27.370]  But that's not our focus right now.
[16:27.530 --> 16:28.310]  That's not what we're going to cover.
[16:29.490 --> 16:34.990]  There are a number of different kinds of surveillance systems that purport to be lifesavers of one kind or another.
[16:35.270 --> 16:36.910]  And many of them really actually are.
[16:37.610 --> 16:46.230]  A great example of this is medical monitoring systems that medical tech companies are building and that are being implanted right now.
[16:46.290 --> 16:47.630]  Especially things like pacemakers.
[16:47.630 --> 16:57.070]  Just getting full-time EKG data, even if it's batch offloaded, can be really, really useful in treating medical conditions.
[16:57.410 --> 17:11.630]  On the other hand, if you have an implant which is monitoring what your heart is doing in real time and can alert doctors and paramedics, as soon as something starts going wrong, you have a huge, huge advantage to survival.
[17:13.130 --> 17:21.730]  Another example along similar lines are fall detection systems that are used for elderly people or for solitary industrial workers.
[17:21.930 --> 17:26.210]  If you're out in the factory floor, you're working alone, you're working swing shift, something like that.
[17:26.390 --> 17:30.070]  You fall into something, nobody sees you because the factory is half empty.
[17:30.970 --> 17:34.790]  You know, the system tells someone who can then come and make sure you're okay.
[17:36.370 --> 17:45.450]  Another example are the remote diagnosis and reconfiguration systems that happen that are being put into luxury automobiles.
[17:45.830 --> 17:53.150]  If the car gets into an accident, a lot of the times these systems can call the police and call paramedics, even in the event that everyone in the car is unconscious.
[17:53.510 --> 17:58.710]  And like, you know, you run into a tree on a dark road at night and the police show up anyway.
[17:59.950 --> 18:03.110]  So manufacturers are packaging these systems into luxury automobiles.
[18:03.110 --> 18:11.390]  They're marketing them to large enterprises or they'll require very good insurance or private medical care in countries which have public insurance to get them.
[18:11.590 --> 18:13.230]  These are luxury surveillance.
[18:14.870 --> 18:20.050]  One of the largest categories of modern surveillance is surveillance that's done for marketing purposes.
[18:21.430 --> 18:22.790]  It's an interesting area.
[18:22.930 --> 18:25.270]  Is this good for the person that's being marketed to?
[18:25.990 --> 18:37.870]  To some extent, corporations do market profiling so they can manipulate subjects into spending money on things that maybe they don't really want to spend money on, you know, regardless of what their private desires or their prior desires are.
[18:38.330 --> 18:43.650]  But there's also an element, however, of showing the subject the things that they're most likely to want.
[18:43.870 --> 18:50.350]  To some degree, you can actually see this as kind of a service, ensuring that people are aware of things that meet their desires.
[18:51.130 --> 18:57.630]  Because that advertising is also trying to manipulate and create that desire, this is a much more kind of muddy category.
[18:58.670 --> 19:01.530]  But really, that communication is not a one-way street.
[19:01.850 --> 19:08.550]  This becomes much more obvious when you look at market profiling that was done in more and more high-end products.
[19:08.770 --> 19:17.410]  As you go further up the scale in something like, you know, ridiculously expensive cars or even more so, ridiculously expensive private airplanes.
[19:17.650 --> 19:20.370]  There's a very small pool of buyers for these things.
[19:20.910 --> 19:31.910]  And a lot of the time, it becomes much more of a conversation where a company like Boeing is saying, or Cessna even more so, is saying to the people, hey, what do you want in your next jet?
[19:32.370 --> 19:34.770]  You know, and getting this kind of direct feedback.
[19:35.230 --> 19:42.010]  You know, and so there's this blurry area where surveillance becomes a conversation, becomes more of a two-way street.
[19:44.570 --> 19:49.750]  Even once data has been captured by a system, it's not necessarily treated the same for everyone.
[19:50.090 --> 19:58.830]  A lot of record systems literally have features explicitly designed to protect some people's privacy to a greater degree than others.
[19:59.230 --> 20:06.190]  We know this, among other reasons, because people are regularly fired for looking at people's data who they shouldn't be looking at.
[20:07.130 --> 20:13.630]  You know, they're looking at the medical records or the tax returns or sealed arrest records of the rich and famous.
[20:13.970 --> 20:19.590]  In the case of private organizations, this isn't a big deal, or as big of a deal.
[20:19.830 --> 20:24.970]  The rich are in a position to simply demand this kind of treatment and get it.
[20:25.510 --> 20:36.910]  But, you know, and for a lot of them, if they have somebody on staff who just goes around and hunts down information about them that they don't want to be public and have it sealed, that may be a perfectly reasonable investment for them.
[20:37.210 --> 20:40.650]  On the other hand, for state records, that's a much more problematic issue.
[20:40.790 --> 20:43.070]  The state should be blind to these kind of things.
[20:43.290 --> 20:48.410]  That's not an acceptable position for a state that takes a fundamental position of citizen equality.
[20:51.150 --> 20:59.650]  One of the more common forms of non-marketing economic surveillance is fraud detection for credit and debit cards and checks and balance transfers.
[21:00.190 --> 21:09.810]  This is another form of surveillance that's explicitly working in favor of the user because the user's interests are mostly aligned in this case with those of the bank.
[21:10.190 --> 21:14.590]  Of course, to be eligible for this kind of surveillance, you have to have a credit card and use it.
[21:14.730 --> 21:16.230]  You have to write checks.
[21:16.230 --> 21:23.490]  You have to spend money in ways that can be monitored, which generally means you are moving further up the economic scale.
[21:23.490 --> 21:29.550]  If you have the money, you can even pay extra money to be more closely watched.
[21:29.910 --> 21:38.290]  You know, you can have, oh, let's, you know, get this extra fraud alert for 50 bucks a year, whatever it is, or the identity theft protection system.
[21:40.230 --> 21:48.670]  So there are a few categories of surveillance that seem to help everyone fairly equally, like environmental surveillance and especially stuff like air and water quality monitoring.
[21:49.350 --> 21:55.510]  However, even these kinds of systems aren't necessarily that equally distributed.
[21:58.150 --> 22:00.870]  Comprehensive environmental quality testing is really expensive.
[22:01.830 --> 22:04.030]  Different cities can afford it at different levels.
[22:04.530 --> 22:08.310]  Inside a city, different areas receive different amounts of attention.
[22:10.090 --> 22:14.490]  Some, you know, some areas, well, let's see, you know, we've got a whole bunch of high rises here.
[22:14.490 --> 22:20.570]  Yeah, we can install one, you know, one water quality monitoring system and protect a whole bunch of people.
[22:20.890 --> 22:23.370]  Well, let's look at property values in that area.
[22:23.650 --> 22:27.150]  You know, there may be a correlation in some cases, in a lot of cases.
[22:29.150 --> 22:37.730]  This disparity gets even larger in cases where private citizens decide to supplement the public monitoring that happens.
[22:38.630 --> 22:42.370]  Home lead and asbestos testing, radon testing, that kind of thing.
[22:42.550 --> 22:44.210]  This happens, like, regularly.
[22:44.550 --> 22:48.130]  And even stuff like water and air quality, it happens occasionally.
[22:49.490 --> 22:55.590]  The same logic applies to traffic counting systems that are used to measure road repair and determine what needs funding.
[22:55.910 --> 23:01.770]  In theory, the transit agency should distribute these sensors fairly across an entire transit district.
[23:01.770 --> 23:04.290]  But the system is subject to social pressure.
[23:04.510 --> 23:11.190]  It's subject to areas where, oh, this area, people have time to file a whole bunch of complaints when the new pothole shows up.
[23:11.670 --> 23:17.890]  And, you know, so the traffic monitors go out better, you know, go out more frequently, they have a better picture of the traffic.
[23:18.190 --> 23:20.250]  Because there's more data, there's more funding.
[23:21.570 --> 23:31.030]  It's not always the case that this kind of stuff is even implicitly aligned with population interest, though.
[23:31.270 --> 23:39.290]  There was a really interesting case in 2001 where the city of New York actually banned private air quality testing.
[23:39.510 --> 23:43.730]  This was in the aftermath of the 9-11 incident.
[23:43.730 --> 23:51.510]  And the debate was still raging over what the long-term problems were going to be of exposure to the dust from the collapsed buildings.
[23:51.510 --> 23:53.930]  So they banned air quality testing for a while.
[23:54.990 --> 23:57.850]  Because the liability cases were still out.
[24:00.070 --> 24:06.730]  Another category of embedded surveillance is the alarm, or really the entire gamut of personal protection systems.
[24:06.730 --> 24:12.650]  You can have your house, and your car, and your laptop, and your mobile phone, and your person all watched for a fee.
[24:13.370 --> 24:19.630]  How much good these services do you varies, among other things, by how much you want to pay for them.
[24:20.270 --> 24:25.790]  You know, you can get a simple alarm system that'll call the police after your house has been broken into at one price.
[24:25.970 --> 24:28.970]  You can get a silent alarm with cameras at another price.
[24:29.230 --> 24:31.890]  You can get private armed response at a third price.
[24:32.170 --> 24:35.030]  You can get on-site armed guards at another price.
[24:35.910 --> 24:39.150]  Surveillance most strongly benefits the people who pay for it.
[24:40.010 --> 24:42.830]  In fact, you often get kind of a race to the bottom.
[24:43.170 --> 24:48.110]  The one house on a block that doesn't have a private armed response team gets robbed.
[24:48.550 --> 24:50.830]  Everybody else lives in a paramilitary camp.
[24:50.830 --> 25:10.690]  If you go to LA, there are entire stretches in West Hollywood along Sunset Boulevard along some of the very rich neighborhoods where every single house has a cement wall which is topped with broken glass and has a sign at the bottom of the driveway that private goons with guns will show up if you break in.
[25:12.870 --> 25:22.890]  It's only a fairly small fraction of the surveillance in a modern city, but surveillance that's used to stop crime is kind of what we think about when we think of surveillance.
[25:23.310 --> 25:30.530]  You know, you have CCTV cameras on rooftops or overlooking back alleys or places where people congregate, especially young people.
[25:31.050 --> 25:34.470]  You have the Ring of Steel that's being built around the City of London.
[25:35.010 --> 25:39.550]  The same thing that was built and is now being rebuilt around Lower Manhattan.
[25:39.550 --> 25:41.750]  You have airport security scanners.
[25:42.550 --> 25:50.790]  So the efficacy of cameras, especially cameras that aren't monitored live, in actually stopping crime is in severe doubt at best.
[25:51.050 --> 26:02.690]  There have been a bunch of studies, especially an internal study by the London police that was leaked, that have shown that cameras have little or frequently no effect whatsoever on stopping crime.
[26:02.690 --> 26:09.730]  In fact, the only crime that the City of London study showed that cameras actually prevented was public urination.
[26:10.450 --> 26:11.290]  That's it.
[26:11.730 --> 26:13.830]  That actually had an effect, nothing else.
[26:14.910 --> 26:24.770]  So police and municipalities are now turning this argument around and saying that, oh, we need to install smarter, better connected cameras that can be monitored live.
[26:24.770 --> 26:28.370]  A form of so-called softer network infrastructure.
[26:28.750 --> 26:31.610]  And that'll theoretically allow the police to intervene in real time.
[26:32.210 --> 26:37.190]  Of course, this means hundreds and hundreds of millions of dollars of more expense per installation.
[26:37.570 --> 26:40.890]  I think the number for Lower Manhattan was $224 million.
[26:41.870 --> 26:46.490]  Assuming it comes in on time and on budget, it'll probably be three times that.
[26:46.490 --> 26:53.790]  And there's no evidence that this will be any more effective than cameras that aren't recorded live or that aren't viewed live are.
[26:54.650 --> 27:03.670]  Especially not in comparison to spending $750 million or whatever it ends up being on more police presence on the street where they can actually do something actively.
[27:03.970 --> 27:08.790]  Or all of the traditional police techniques that we know actually work and that have a very long record.
[27:09.950 --> 27:15.950]  So regardless of their utility, what's interesting from our perspective is the social division that these cameras create.
[27:15.950 --> 27:18.550]  Or rather the one that they reproduce and sort of highlight.
[27:18.870 --> 27:22.850]  The camera creates three explicit social roles in its interaction.
[27:23.130 --> 27:25.190]  The watcher, the watched, and the bystander.
[27:25.730 --> 27:29.530]  The distinction between the last two categories is especially interesting.
[27:29.970 --> 27:35.250]  A camera installed for security is not intended to catalog all actions in this space.
[27:35.610 --> 27:42.210]  It's only intended to catalog the set of actions that the watcher deems interesting.
[27:42.830 --> 27:48.130]  Cameras thus select out a subset of actions and a subset of the people that pass through their frame.
[27:48.130 --> 27:52.430]  The actual or just the stereotypical committers of those actions.
[27:52.710 --> 27:58.910]  This is reflected in both the attention that the watcher pays to the video and in the positioning of the camera.
[27:58.910 --> 28:14.690]  When cameras are intended to prevent crimes against people, the bystanders, the people who are in the, you know, the class that aren't the expected offenders, become the protected class, giving an even more stark opposition.
[28:14.690 --> 28:19.070]  For the watched, the cameras represent hostility, they represent suspicion.
[28:19.390 --> 28:22.790]  For the bystanders, they represent security and freedom from trouble.
[28:23.130 --> 28:31.230]  For the watcher, they represent the ability to choose who's in what class, which actions get responded to, and which actions get ignored.
[28:32.990 --> 28:40.910]  This is especially relevant in the case where you have police-run cameras observing police actions.
[28:40.910 --> 28:44.570]  The police in the frame should belong to the protected class.
[28:44.810 --> 28:49.150]  You know, they are one of the set of people who the cameras should be protecting.
[28:49.450 --> 28:56.910]  But by fraternity with their fellow police, if nothing else, sometimes even by policy, they're in the watching class.
[28:56.910 --> 29:02.090]  They're in the class that get to partially decide what actions are responded to.
[29:02.090 --> 29:10.010]  This is demonstrated nowhere better than on occasions when police brutality should have been captured by police cameras.
[29:10.250 --> 29:13.890]  But when there's a trial, oh, we don't know what happened to that footage.
[29:14.310 --> 29:14.830]  Sorry.
[29:15.350 --> 29:26.530]  So, another interesting thing we should note, without exception, the protected are likely to belong to a higher socioeconomic stratum.
[29:26.530 --> 29:38.290]  Not only will the well-off be almost always perceived as less likely to be the offenders in any given situation, they will also spend more time in watched areas.
[29:38.590 --> 29:44.410]  It is no accident that the most heavily watched areas in the world are the homes of the global financial markets.
[29:44.410 --> 29:56.050]  You know, there are even cases where there are setups that allow the privilege to avoid the scrutiny entirely for themselves, but still benefit from it.
[29:56.930 --> 30:01.070]  There's, I mean, the classic, classic example is diplomatic privilege.
[30:01.330 --> 30:03.970]  You know, this is the ultimate form of bypassing surveillance.
[30:04.210 --> 30:06.190]  Of course you cannot search my possession.
[30:06.450 --> 30:07.610]  I am a diplomat.
[30:08.370 --> 30:16.950]  There's the clear system, which is now defunct and may be coming back again at some point, which, if you bought into it, allowed you to bypass certain parts of airport security.
[30:17.910 --> 30:24.410]  Or you can just get driven into your office building in a private limo driven by an off-duty cop with dark tinted windows.
[30:24.710 --> 30:27.070]  You know, and pay to have the plates changed every now and then.
[30:29.010 --> 30:37.670]  Another interesting note in this category is the collusion of surveillance technology and restricting speech in public private areas like malls.
[30:38.050 --> 30:52.330]  So, in those cases, the division isn't simply between the social categories of, you know, who the likely offenders are, but between those who are willing to remain silent or who agree with management.
[30:52.330 --> 31:00.370]  And those who, you know, those who actually want to speak up and say something.
[31:01.810 --> 31:06.170]  So, we've seen kind of a consistent pattern here in the effects of surveillance.
[31:07.150 --> 31:16.230]  You know, surveillance can be good and bad, but it's an activity which directly embodies the power structures of the societies that perform it.
[31:16.230 --> 31:29.750]  The technologies that implement surveillance render these social structures very literally in steel and glass and silicon and embed it into the fabric of the city in a, you know, in a very literal sense.
[31:29.910 --> 31:37.450]  You know, these things are being built into buildings, into subways, into streets, in more and more ubiquitous ways.
[31:38.230 --> 31:52.670]  So, we have a consistent pattern where people with money or power or social status can either act less to be watched in bad ways and more in good ways or simply enjoy greater protection as a secondary effect of their social standing.
[31:52.950 --> 32:04.740]  It's just sort of another reflection of the ubiquitous privilege that they move through, often completely unaware that everyone doesn't experience the same kind of surveillance profile.
[32:05.520 --> 32:07.460]  This shows no signs of stopping.
[32:08.540 --> 32:21.900]  On the contrary, there's a whole class of initiatives called smart city initiatives, which, among other things, are looking at pushing things like environmental monitoring, a lot of the kinds of good surveillance, out to portable devices.
[32:21.900 --> 32:25.840]  You have like, oh, your smartphone can be an air quality monitor, that kind of thing.
[32:26.380 --> 32:34.060]  You know, this may be good in the end, but in the meantime, the gadgets that are enabling this kind of thing are luxury devices.
[32:34.060 --> 32:37.960]  I mean, I'm sure a lot of people in the room have them, but they're not free.
[32:38.140 --> 32:39.900]  They're not even necessarily cheap.
[32:40.360 --> 32:43.840]  And this is reproducing the same kind of cultural divide.
[32:44.980 --> 32:48.360]  You know, it may propagate down in the end, but it's not going to happen immediately.
[32:51.240 --> 32:52.940]  So, what can we do about all of this?
[32:53.720 --> 32:59.860]  Both the growing surveillance state and the inequality in surveillance need to be addressed kind of in equal measure.
[33:00.540 --> 33:04.200]  Hackers and makers are in a really interesting and unique place here.
[33:04.440 --> 33:12.220]  We have both the technical background in the systems and often the socio-political understanding to see the full breadth of the problem.
[33:12.520 --> 33:20.120]  We also have at least a somewhat organized community that can work together, albeit in a very sort of distributed way.
[33:21.420 --> 33:27.080]  We're in some ways a very privileged group, but in other ways we're a frequently fairly marginalized group.
[33:27.600 --> 33:33.560]  This puts us in a position of, if we try, being able to reach out kind of in both directions.
[33:35.100 --> 33:37.800]  So, there are four things that we can do, roughly speaking.
[33:38.000 --> 33:39.040]  We can document the problem.
[33:39.180 --> 33:40.640]  We can raise awareness of it.
[33:40.720 --> 33:43.860]  We can work to change existing projects that are being implemented directly.
[33:43.860 --> 33:45.720]  And we can subvert these systems.
[33:46.060 --> 33:52.440]  In a lot of ways, this is the same catalog of tactics that's true for any kind of social change, any kind of social movement.
[33:54.160 --> 33:59.720]  And like any other social change movement, there are existing groups working in a lot of these areas.
[33:59.840 --> 34:02.460]  If you work with these groups, you can have more of an effect.
[34:03.920 --> 34:12.480]  So, the details of the ways that different groups watch us and what happens to that data are not often public.
[34:12.680 --> 34:15.640]  And they're certainly not aggregated and actively publicized.
[34:16.340 --> 34:22.320]  Helping to find and collate this kind of information is a basic first step for any kind of other work.
[34:22.420 --> 34:30.300]  You can't, you know, try to talk to people about this before you understand the scope of the problem and how the system is working.
[34:30.300 --> 34:36.000]  This can be as simple as doing things like marking and updating maps of where surveillance cameras are in neighborhoods.
[34:36.520 --> 34:38.520]  On the other hand, it can be a lot more involved.
[34:39.200 --> 34:43.880]  It can mean doing background research on the technical systems that are used for surveillance.
[34:44.220 --> 34:53.800]  It can mean filing Freedom of Information Act or whatever the local equivalents are in other countries to find out what's going on with these systems.
[34:54.040 --> 34:59.660]  It can mean reverse engineering, you know, physical examples of the surveillance systems themselves and documenting them.
[35:00.360 --> 35:04.540]  There was an interesting case of this that happened recently in Seattle.
[35:04.800 --> 35:07.400]  There was a wrongful arrest case involving a hacker.
[35:07.700 --> 35:11.760]  And the police weren't forthcoming of video evidence that was relevant to the case.
[35:11.960 --> 35:13.740]  You know, oh, it got deleted.
[35:15.780 --> 35:19.720]  So, the guy in question wasn't really willing to take that answer.
[35:19.900 --> 35:27.700]  So, he determined that the surveillance management system that the police were using kept secure logs any time media was deleted.
[35:28.040 --> 35:30.260]  And he had his lawyers request the logs.
[35:30.540 --> 35:36.460]  You know, clearly the police should be able to back up their supposition that this data was deleted with the logs.
[35:36.460 --> 35:45.080]  So, he analyzed the log files, figured out how they marked data and how they tracked it, and proved that the video did still exist.
[35:45.340 --> 35:50.320]  It was, you know, this video can be deleted after 90 days, but it's kept on a space available basis.
[35:51.180 --> 35:55.000]  So, then he went back to the police and said, well, no, this video still exists.
[35:55.140 --> 35:55.740]  Here's the proof.
[35:56.240 --> 35:57.200]  Turn over the video data.
[35:57.460 --> 35:58.040]  They did.
[35:58.680 --> 36:07.140]  So, now this means, and this same, you know, most surveillance video management systems used by the police do the same thing.
[36:07.640 --> 36:13.960]  Now, lawyers who are in the same position with their clients can take the same action.
[36:14.100 --> 36:19.780]  They can request the logs, have somebody analyze those logs, and see if the video is actually gone.
[36:20.120 --> 36:28.660]  And also, because the surveillance systems are actually somewhat decently designed, there, it shows the difference if data is ever manually deleted.
[36:29.860 --> 36:33.400]  So, if the police want to go and say, oh, well, that data was deleted.
[36:33.700 --> 36:36.520]  Well, no, that data was deleted by this user intentionally.
[36:37.100 --> 36:44.700]  You know, that's a very different position in court, and a judge can take that into account, even if they can't find out why the data was deleted or get it back.
[36:49.120 --> 37:00.740]  So, it's not just enough for us to understand these issues any more than it is for the knowledge of these issues to be fragmented and not really centrally brought together as a cohesive image.
[37:00.860 --> 37:02.820]  The rest of the world needs to understand this, too.
[37:03.120 --> 37:09.180]  We've been telling people about, obviously, computer security and also online privacy for a really long time.
[37:09.520 --> 37:22.260]  But we need to start speaking up more and talking to the larger community and saying, hey, all of these issues that we've been talking about in terms of online security, yeah, these happen at your grocery store now, too.
[37:22.260 --> 37:24.440]  You need to deal with them there, as well.
[37:24.860 --> 37:28.480]  You need to deal with this, you know, you need to start thinking about the same kinds of actions.
[37:31.500 --> 37:38.480]  You know, and it's important to get through the message, as well, that this kind of surveillance doesn't necessarily help you.
[37:38.720 --> 37:41.560]  Just because you're on video doesn't make you safer.
[37:43.500 --> 37:51.900]  Most new surveillance projects and the legal structures that new surveillance projects are built on aren't created in a total vacuum, which gives us a place to act.
[37:52.480 --> 38:05.340]  This is the kind of the usual stuff of civic activism, whether that means protest, or contact with lawmakers, or raising the sort of surveillance at planning meetings, getting into all the annoying technical stuff, whatever.
[38:06.540 --> 38:11.720]  One of the really interesting issues here is that modern surveillance exists kind of at a couple of levels.
[38:11.720 --> 38:19.680]  It can exist at, like, the national law level and at sort of the Internet level, but it also is very specifically geographically located.
[38:19.960 --> 38:26.100]  You know, surveillance cameras may be plugged into the cloud, but they exist in one specific physical location.
[38:26.280 --> 38:28.060]  And you have to act on both sides.
[38:28.360 --> 38:30.640]  You know, there's both the, well, what if the U.S. had national privacy protection laws like Europe does?
[38:33.480 --> 38:55.040]  And then there's the, hey, you know, let's go talk to the businesses that we interact with and say, down to the guy at the bodega, hey, you realize that that camera isn't going to stop somebody from robbing your store, but it does mean that those video records can be subpoenaed if the police think you're dealing drugs and want to harass you.
[38:57.800 --> 39:00.760]  You know, so you have to kind of work at both levels like that.
[39:02.020 --> 39:03.840]  Subverting systems is something we're really good at.
[39:03.980 --> 39:04.800]  It's kind of what we do.
[39:05.540 --> 39:07.180]  It can take a lot of different forms.
[39:07.320 --> 39:11.260]  You can have pranks that just show how ridiculous this whole surveillance culture is.
[39:11.900 --> 39:19.660]  But it can also be practical tools and techniques for either avoiding surveillance in the problem or just for countering the power dynamic that it creates.
[39:20.200 --> 39:27.640]  You've got things like the Tor project that Jake was talking about in the other room for proxying content and fighting traffic analysis.
[39:27.640 --> 39:37.940]  And you've got OTR for secure messaging that both exist for getting around the first problem of dealing with avoiding surveillance.
[39:38.800 --> 39:47.480]  But in the last part of this talk, I'm going to talk about a new competition for projects in the second category to help equalize the power balance in terms of surveillance.
[39:49.740 --> 39:55.600]  So, the greatest power imbalance in the sphere of surveillance exists between the private citizen and the state.
[39:55.780 --> 40:00.180]  And this is most exacerbated when the citizen is trying to take action against state violence.
[40:00.520 --> 40:03.840]  Police and military brutality is real and it happens all the time.
[40:04.020 --> 40:07.080]  And it happens all over the world in every country.
[40:07.440 --> 40:10.100]  The worst excesses happen when no one is watching.
[40:10.540 --> 40:17.820]  Even though surveillance tools are widely deployed throughout the world, there are many countries where surveillance is only ever going to be a tool of oppression.
[40:17.820 --> 40:19.720]  Where it's never going to work for you.
[40:20.380 --> 40:24.380]  In a lot of situations, there's no question whether or not you're going to show up on video.
[40:24.700 --> 40:28.220]  And that if that video can be used to incriminate you, it will be.
[40:29.180 --> 40:32.340]  This is especially true for large political demonstrations.
[40:32.980 --> 40:48.420]  However, if the state decides to suppress a demonstration, especially if they do decide that, yeah, we're just going to send in the tanks and, you know, shoot everything that moves or, you know, whatever, especially in the more brutal suppressions, they often suppress the video.
[40:48.640 --> 40:50.860]  They don't want anybody seeing them doing this.
[40:51.960 --> 41:01.000]  It's questionable at best if getting video out from places like Burma or Iran can actually materially affect the situation on the ground in an immediate sense.
[41:01.000 --> 41:04.860]  But on the other hand, it does affect public opinion in the long term.
[41:05.220 --> 41:15.820]  You know, if you look at the Naina case in Iran, that was a huge, huge organizing point that really prolonged those protests and really prolonged public support.
[41:16.480 --> 41:18.180]  Because that video got out.
[41:19.740 --> 41:24.500]  Even more interesting to us are cases where there is recourse to the rule of law.
[41:24.680 --> 41:30.060]  You know, that you can actually take action after the fact, but there's still significant corruption.
[41:30.060 --> 41:32.600]  You know, the police are still going in there and beating people up.
[41:33.040 --> 41:36.860]  This is a really big part of the world, both the developing world and the developed world.
[41:37.720 --> 41:40.680]  So the obvious way to fix this is to just shoot your own video.
[41:41.380 --> 41:43.620]  You can't install a city-wide camera network.
[41:43.700 --> 41:44.380]  It's too expensive.
[41:44.400 --> 41:46.360]  It's not a reasonable thing to do, but you don't have to.
[41:46.940 --> 41:50.940]  All you need is a few cameras that are in the right place at the right time.
[41:51.880 --> 41:56.680]  You can use a phone that shoots video or a cheap, you know, handheld camera.
[41:57.040 --> 41:58.920]  But there are four problems with them.
[41:59.220 --> 42:08.640]  First, do you really want to be the guy standing in front of a group of riot cops holding up the video camera, watching them beat the crap out of some guy when there's 200 of them and one of you?
[42:09.600 --> 42:10.660]  No, you don't.
[42:12.900 --> 42:17.500]  Second, if you can't get their badge numbers down, it's a lot harder to take action afterwards.
[42:17.860 --> 42:23.020]  When you're, like, waving your camera, running away from those 200 riot cops, you're not recording anything useful.
[42:23.580 --> 42:26.620]  I mean, it might be useful for PR, but it's not useful in court.
[42:28.660 --> 42:31.280]  The video is stuck on your camera until it gets uploaded.
[42:31.700 --> 42:39.960]  When the riot cops catch up to you and take you in, even if they didn't see you holding up the camera, they're still going to beat the crap out of you when they look on your phone.
[42:40.360 --> 42:43.420]  And if you don't let them see your phone, they'll just beat the crap out of you until you do.
[42:46.900 --> 42:50.080]  Fourth, there's no way to collate this video or manage it online.
[42:50.400 --> 42:52.220]  Yeah, you can put those videos on YouTube.
[42:52.640 --> 42:58.000]  But what if you happen to be in an area where the cameras didn't get a good look at everybody's face?
[42:58.360 --> 43:05.540]  Do you really want to, you know, or you got a much better look, or you recorded a disempotive conversation that lets somebody identify something?
[43:05.820 --> 43:10.780]  You don't necessarily want that video going out public right away because that can get people killed.
[43:13.500 --> 43:18.260]  Every jurisdiction also has different rules around what you can record and what you can't record.
[43:18.580 --> 43:28.440]  If you have a system that's capable of recording audio and you use it to record stuff on the street, even if you're not recording audio, you can still sometimes get arrested for wiretap.
[43:29.460 --> 43:37.580]  So if you're building a system for the U.S., you'd really like it to not be able to record audio because then you can just tell the cops, no, no, no microphone.
[43:38.120 --> 43:44.360]  You know, they may arrest you anyway, but it's a lot harder, and it's much, much harder to prove that that's wrongful arrest after the fact.
[43:45.000 --> 43:51.900]  On the other hand, you know, 50% of the world has a cell phone, and a lot of those phones are going to be recording video in a few years.
[43:52.860 --> 44:01.260]  So in the interest of kind of leveling the playing field a bit, I want to announce a competition to build deployable video cameras.
[44:01.540 --> 44:09.180]  A combination of off-the-shelf components, open hardware, and free software that can be deployed in an urban environment.
[44:09.720 --> 44:11.120]  There are three categories.
[44:11.640 --> 44:20.480]  Aerial cameras, static cameras, and software-only solutions to run on existing mobiles that provide discreet live uploading and appropriate media management.
[44:21.140 --> 44:37.140]  So for UAV, I mean, maybe somebody will build a fixed-wing, you know, actually flying option, but much more likely it's going to be like the quadcopter that perches on the roof and can sit and stare, you know, and it can get up there, and it's still a fairly cheap platform,
[44:37.200 --> 44:39.000]  but it can go places nothing else can.
[44:39.560 --> 44:47.580]  Or it's the camera that, you know, you throw over a phone line like a pair of shoes, and it hangs out up there, and so you tell it to just drop and pick it up later.
[44:48.860 --> 44:57.560]  The goal of the competition is to spur work on these kind of devices and similar ones as much as it is to develop immediately fieldable solutions.
[44:58.780 --> 45:10.340]  It's probably going to take a few iterations to get to something which is really going to be actively useful in the field and is going to be cheap enough, but on the other hand, if you don't start, you'll never get there.
[45:11.480 --> 45:13.280]  The competition is open to everyone.
[45:13.840 --> 45:16.780]  Hackerspaces and makerspaces are especially encouraged to enter.
[45:16.780 --> 45:18.920]  This is a relatively big problem.
[45:19.520 --> 45:21.960]  Coming up with a good solution is going to require teamwork.
[45:22.840 --> 45:38.720]  The only firm requirements are that everything that's developed for the competition has to be released under open-source software and open hardware and open documentation licenses so that it can be actually reproduced by the groups in question without needing some kind of centralized supply chain,
[45:38.840 --> 45:42.500]  and that the system needs to be legal in some kind of chosen jurisdiction.
[45:43.180 --> 45:51.060]  If you're in the U.S., there's no point in submitting a system that uses 3 GHz radios because that's just going to get somebody arrested that much faster.
[45:53.620 --> 45:55.800]  So, this is the set of judging criteria.
[45:56.240 --> 45:58.020]  I'm not going to go through all of these.
[45:58.500 --> 46:01.260]  I'm not going to go into the details and scoring and that kind of thing right now.
[46:01.660 --> 46:05.680]  There's a competition website that has all of this information available on it.
[46:07.460 --> 46:10.320]  Two notes, and I'll just give the URL right now so you have some time.
[46:11.940 --> 46:25.280]  The cost is definitely a criteria for this competition, but it's expected that the first generation of devices are probably going to be too expensive by, you know, at least an order of magnitude for what we'd really like to see for the field.
[46:25.660 --> 46:27.520]  But, you know, that happens.
[46:27.700 --> 46:28.940]  The prices will come down.
[46:31.480 --> 46:38.940]  Second, teams are really strongly encouraged to get together with groups that might actually feel these things in the real world.
[46:39.360 --> 46:44.860]  Talk to these people, work with them, you know, do use case studies, find out what they think would be useful.
[46:46.000 --> 46:50.400]  If you work with a group, both working with the group will be taken into account.
[46:50.540 --> 47:02.540]  And also, if you get a group that says, hey, you know, we organize protests all the time, like Copwatch in Berkeley that has a bunch of other branches, you know, and this would be exactly the camera that we need.
[47:03.160 --> 47:03.560]  Great.
[47:03.800 --> 47:06.580]  You know, and that'll also get taken into account for the judging.
[47:07.020 --> 47:11.960]  There's going to be a preliminary evaluation of progress for all the teams that are registered in about six months.
[47:12.580 --> 47:16.640]  I'm hoping to maybe talk about this at the CCC Congress in December.
[47:16.900 --> 47:18.480]  If we have some results then.
[47:19.540 --> 47:25.760]  Basically, if we have about, you know, 75% of the teams that have registered, we've got some real fieldable systems in.
[47:25.920 --> 47:27.920]  We'll do the judging round then.
[47:28.240 --> 47:33.920]  If the development cycle ends up taking longer, you know, we'll talk to the teams and see what the deadlines sound like they should be.
[47:36.920 --> 47:42.040]  So, there's the URL, sldrc.com slash projects slash deployable.
[47:44.020 --> 47:45.120]  So, any questions?
[47:46.320 --> 47:48.160]  And, oh, the slides are at that other URL.
[47:48.640 --> 47:51.060]  So, I'll just leave that up there for a minute for people to write it down.
[48:04.220 --> 48:05.580]  Is there a mic for questions?
[48:06.360 --> 48:06.720]  Okay.
[48:23.590 --> 48:28.930]  So, I actually haven't got... the question was about the New York City air quality testing issue.
[48:29.110 --> 48:32.450]  I haven't done as much research on that one as I should have.
[48:33.290 --> 48:40.730]  There was a professor at Columbia who does environmental systems work in the architecture department who passed that on to me.
[48:40.850 --> 48:42.750]  I think it was a short-term thing.
[48:43.450 --> 48:53.050]  Certainly because their position on the health risks there changed pretty rapidly after it was shown that, yeah, this was, you know, obviously had real health effects.
[48:53.690 --> 48:56.210]  So, I think it was at best a short-term thing.
[48:56.890 --> 49:00.230]  And again, I'm not sure if they ever managed to successfully enforce it or not.
[49:00.350 --> 49:03.670]  But they did actually get to the point of trying to prevent it legally.
[49:07.940 --> 49:18.980]  Do you have any kind of an outline put together so someone could, say, search for local city council meeting minutes, go through steps, bring it to the intentions?
[49:19.160 --> 49:20.520]  Because sometimes they have open forums.
[49:20.720 --> 49:21.120]  Yeah.
[49:22.060 --> 49:26.710]  Do you have something we should be putting together as well?
[49:27.270 --> 49:32.990]  The question was about an outline of, like, questions I'm assuming to talk to city councils with and that kind of thing.
[49:33.530 --> 49:35.370]  I haven't prepared anything like that.
[49:35.370 --> 49:37.670]  I'd be happy to work with someone to do that.
[49:38.850 --> 49:46.590]  The slides of this talk and the script for this talk are available online as far as if you want to get more information about some of the stuff I've referenced.
[49:47.150 --> 49:48.470]  And then, yeah, get in touch with me.
[49:49.430 --> 49:52.610]  I'd definitely be interested in working with people to put that kind of document together.
[49:59.050 --> 50:01.490]  No, I'm probably going to put it up in Subway at some point.
[50:02.290 --> 50:02.890]  But...
[50:02.890 --> 50:03.790]  But,
[50:06.800 --> 50:06.800]  uh...
[50:06.800 --> 50:25.640]  Your contest, uh, stressed me as it's too much so, reminds me of many of the, uh, military department, private, local, local, local, local, local, local, local, local.
[50:25.640 --> 50:42.180]  And so, I see the opportunity for your contest, uh, being used by, not necessarily, John Taylor's citizen, which is a popular revision of Google.
[50:42.620 --> 50:55.710]  It's been used by, uh, organizations that could possibly end a great deal more on, uh, and then when you develop, um, that your citizens will move.
[51:00.000 --> 51:08.500]  So, the question was about, um, who else might end up using the, uh, the results of this competition other than the people who would like to.
[51:08.800 --> 51:12.340]  I guess, here's the thing, um...
[51:12.820 --> 51:16.180]  Police agencies don't have to go to open hardware for their surveillance needs.
[51:16.320 --> 51:17.420]  They just bolt up cameras.
[51:17.640 --> 51:19.340]  And they can afford to put cameras everywhere.
[51:20.220 --> 51:25.620]  And, um, the military and the intelligence agencies are already building...
[51:25.620 --> 51:28.180]  I mean, they already have Predator drones.
[51:28.300 --> 51:30.240]  They don't need to build open-source Predator drones.
[51:30.560 --> 51:45.080]  Um, basically, this is about technology equalization and getting the technology out of just the hands of government agencies and larger corporations that can afford the existing, uh, military-industrial versions.
[51:45.300 --> 51:48.600]  And saying, okay, well, we get to have those toys to play, too.
[52:03.680 --> 52:07.480]  The question was about why make it fit in a legal... legally into a jurisdiction.
[52:08.040 --> 52:10.540]  So, basically, the idea is that, okay...
[52:10.540 --> 52:12.300]  So, if the only law that you're violating...
[52:12.300 --> 52:18.760]  For instance, a couple of states have recently, um, passed or are trying to pass laws as banning, you know, taking pictures of the police.
[52:19.160 --> 52:23.160]  If you break a law that bans taking pictures of the police, that's fine.
[52:23.740 --> 52:33.860]  But, if you ban a law that says, you know, if you're, like, breaking frequency allocation stuff, that kind of thing, that gives them another lever to use to ban these systems with.
[52:33.860 --> 52:42.700]  And then they can say, oh, this is a public safety issue that these, you know, these crazy people are building these systems that don't respect radio allocations or whatever and that kind of thing.
[52:42.840 --> 52:47.940]  And it makes it much, much harder for a legitimate group to actually field a system like this.
[52:48.380 --> 52:56.080]  Um, you know, it's almost certain that if we get these systems being actively used on a regular basis, yeah, of course, they're gonna get banned.
[52:56.460 --> 53:07.080]  But, that's a different category of law versus something around, um, you know, some, you know, existing public safety, existing spectrum laws, existing, you know, that kind of thing.
[53:07.740 --> 53:16.000]  Um, you know, and it's basically saying, you know, can assist, can a, uh, can a group field something like this and not get arrested on completely unrelated charges.
[53:17.660 --> 53:22.660]  Uh, technology is an extension of existing social power structures.
[53:22.860 --> 53:23.280]  Yes.
[53:23.460 --> 53:28.750]  So, it's not particularly surprising.
[53:29.210 --> 53:36.350]  You seem to wish to empower justice in an ideal sense, not just in the particular justice system.
[53:37.010 --> 53:44.950]  What do you see socially taking the initiative to expand their own power using surveillance technology?
[53:46.470 --> 53:53.510]  So, um, the, the question was about who I see taking, you know, taking the initiative to use these kinds of tools.
[53:54.570 --> 54:01.710]  It's already becoming more of a done thing for, you know, okay, so, here's an example.
[54:02.130 --> 54:16.830]  Um, a number of years ago, a friend of mine was involved in some student organizing, organizing work in the UK, um, at Brighton University when the students were doing a bunch of takeovers of, um, administration buildings as part of student general strikes.
[54:17.190 --> 54:24.470]  And his job was to stand in frame and be completely innocuous and do absolutely nothing that could possibly get himself arrested.
[54:25.490 --> 54:29.950]  Because, then, he could go, um, subpoena for those videos.
[54:30.170 --> 54:39.270]  He could go file a Freedom of Information Act request for those videos without there being any risk of incriminating himself in the process of doing so and make sure that the videos were released.
[54:39.270 --> 54:47.110]  So, in some kind of very grassroots ways, people are already taking action to equalize the effects of surveillance.
[54:47.530 --> 54:55.550]  Um, and you're getting similar things where, I mean, like, okay, look at how the, um, look at how G20 went in Toronto.
[54:57.110 --> 55:04.710]  The media picture for surveillance, you know, by the organized news networks was a tiny percentage of the video that came out of there.
[55:04.910 --> 55:12.310]  You had 15,000 minutes of video for every, of, you know, citizen video for every minute of network video that got shot.
[55:12.730 --> 55:15.090]  Um, that got uploaded to YouTube or whatever.
[55:15.390 --> 55:16.930]  And, you know, I mean, round numbers.
[55:17.130 --> 55:20.530]  But, um, this kind of thing is already happening.
[55:20.870 --> 55:22.950]  People are just using tools that really suck.
[55:22.950 --> 55:25.730]  So, let's give them tools that don't suck quite as much.
[55:28.210 --> 55:28.890]  Anyone else?
[55:37.440 --> 55:37.840]  Um,
[55:45.150 --> 55:46.270]  feel free to send stuff to me.
[55:46.670 --> 55:48.610]  And, or, you know, and we can, we can talk offline.
[55:52.340 --> 55:56.580]  Are you familiar with the BBC production called The Last Enemy?
[55:57.780 --> 55:58.580]  No, I'm not.
[55:59.100 --> 56:00.560]  You will want to watch it.
[56:00.740 --> 56:00.920]  Okay.
[56:00.920 --> 56:03.400]  Then buy it from WPBH.
[56:03.600 --> 56:04.040]  Yeah, sure.
[56:05.620 --> 56:07.800]  I think there was one more question.
[56:08.780 --> 56:12.060]  Um, coming back to this gentleman's question, which I thought was good.
[56:12.840 --> 56:27.200]  If you acknowledge that, um, some of these devices may become illegal for various, for reasons that could actually be a psychicality, for example, violating a certain frequency spectrum usage guideline.
[56:27.200 --> 56:27.340]  Mm-hmm.
[56:28.220 --> 56:30.940]  Um, and I think perhaps the answer could be a good word.
[56:31.160 --> 56:32.960]  I don't necessarily think they would become illegal.
[56:33.140 --> 56:47.340]  But, um, if they were going to become illegal for a technicality like that and doesn't directly address, like, the, uh, moral reasons you use them or not, um, why worry about that kind of locality up front in this context?
[56:47.340 --> 56:51.560]  They'll be useful before they become illegal and useful after they become illegal.
[56:52.800 --> 57:02.340]  The question was, um, why worry about, uh, systems becoming illegal on technicalities when that happens before the, you know, before they've actually, that's actually been a problem.
[57:02.340 --> 57:14.960]  Um, basically because if these systems get fielded for real and real organizations are using this, um, you don't want to get someone in any more trouble than they may get into anyway.
[57:15.120 --> 57:28.620]  And if that means that in addition to, you know, whatever charges somebody gets laid in, they get, um, also charged with, uh, you know, violate, you know, FCC charges or whatever and some additional fine.
[57:29.220 --> 57:33.860]  That's some real organization that has to pay that fine when we find out that, oops, somebody screwed up.
[57:34.200 --> 57:36.900]  Um, so it seems like it's kind of an aspect of due diligence.
[57:37.160 --> 57:49.460]  You know, it's for the same reason that, okay, if you're going to make an actually autonomous UAV version of this thing, you need to be very careful that it's not going to crash into somebody and kill them because then whoever's fielding it gets charged with,
[57:49.580 --> 57:52.820]  like, obstruction of justice and manslaughter.
[57:53.440 --> 57:56.820]  And you'd really like to not give somebody in the field that additional letter.
[57:59.120 --> 58:01.680]  So, I think we're pretty much at time.
[58:01.960 --> 58:04.280]  Um, if anyone else has questions, you can take a moment.
[58:04.280 --> 58:04.360]  Thank you.