I am a Glass Explorer and keep up to date with all the news and information on Glass. a particular subject which keeps cropping up is that of the use of video data collected through Glass. Although it is true that its capabilities for Big Brother-type activity are present in the device’s technology, it really is no different to your phone in many respects in terms of tracking your whereabouts, what you buy etc. As for privacy, the average city dweller walks past a hundred active CCTV cameras a day so there’s no real change there either- do you ever complain in a store that the cameras are following you? You assume they are using the data they collect for security but in reality, you have no control over where it can be used.
The gathering of information via video in a Big-Brother-fearing manner is another huge big-data analysis challenge that is way beyond the capabilities of our computing networks even into the foreseeable future.
The issue is that to be a great Big Brother, you have to be able to look at all the data and use it somehow. Who will have the computing power required to analyze 100 million users’ personal videos every day for example – never mind the billion tweets and emails that they also generate?
Google Glass Video – 5 hrs per day = 5Gb of data
Let’s stick with video as the other data available from your Android is already being analyzed and is old news.
An average hour of MP4 at 720p is about a Gigabyte. So an average user with 5 hours video per day and 100 million users; as Big Brother you would be processing the data amounting to perhaps (at a very conservative estimate of 500 million Gigabytes (476 Petabytes) per day. If you had this computing power on hand, the genome project could have been completed years ago. As it is, although we know what genes there are, matching them to the creation of over 100,000 proteins is apparently a bitch to compute. The biggest universities in the country have spent over a decade working on it using Supercomputers and complex math. So back to our Google Glass video surveillance by the government (or Google?). 476 Petabytes of data processing would require not only a server farm the size of Greenland to store it and index it all: 576 Petabytes needs 488,000 One-Terabyte hard drives like the one in your laptop to store it on.
Cost? One day’s storage of 476 Petabytes at say $50 per drive is 488,000 x 50 = $24m for the drives plus about the same every week to manage them. That’s just the hard drives for one day’s storage. Add in there the cost of the computers required to run them….they are probably about $1000 per server blade running a max of 10 hard drives in an array so you need 48,800 of those too. The processing required to analyze them all would be another server per drive. That is now half a million servers plus drives. Oh and there needs to be data integrity too so add a 20% overhead to all of that. The actual cost of just the storage and processing hardware is phenomenal. Now add a new satellite or two as this is mobile data and bandwidth would need to be created to cope with the data flow. We are talking many billions of dollars, and thousands of employees to install and maintain such a vast computer.
Can’t get your head around those big numbers? Google processes about 20 Petabytes per day of user data across its servers. 476 Petabytes of user data is going to need 20 Google-sized server farms to process it. The thought that the government can suddenly resource all that (without you knowing too) is unthinkable. The cost of each “Google” sized complex would be hundreds of millions – trillions would be the actual total cost when communications, human resources and power consumption are included.
Pretend they did.
Assume they managed to do this, what would Big Brother get from the many $Billions it would require to become a decent-sized Big Brother operation? Um… a lot of video of people doing their daily stuff. Combing out the faces from that video alone would require such a large server farm, it would be cheaper to hire cops to just go walk the beat instead. And there we have it. Google Glass users will have a choice in all this too. The video cannot be “Switched on” by some nefarious organization as part of some NSA plot. This type of nonsense has been touted before. Why is it so ridiculous a scenario? The user can tell when the video is running. Even without the display being active, the chips on the circuitry inside warm up. You can feel them get warm as the processor works harder. Also, the battery would be used up and noticeably drained for no reason. so all these comments are really sensationalist and not based on any technical observation or conclusion. As a Glass Explorer I will not be in the slightest bit concerned about the video my Glass takes as I cannot see what interest that anyone would have in it other than curiosity. In line with the regular social acceptability of pointing a camera at someone and taking a picture, you risk being punched in the face if you dont choose your subjects carefully or politely. Glass is no exception and I imagine there will be an ER that will be picking a prism display out of someone’s forehead at some point for this reason. Technology changes but we don’t change much. The next generation will assume that a wearable computer is as normal as a telephone with a rotating dial to the generationsof years gone by. Including me.
A more likely use for facial recognition – not Glass though.
Cops will walk their usual beats (probably more of them) and their scanning cameras (which are unlikely to be Google Glass, as integration into their current attire is more likely) and their Facial Recognition Camera will pick out those that they already know are active criminals and those who are wanted already. Hey ho. What is different to how things are now? They use the CCTV in public places to do this anyway, and most of these cops already know the local on-foot criminals on their patch – its their business to know them and keep an eye on them. All a Facial Recog. device will do is enable them to spot their targets easier.
Will they spot the wrong people? Much less likely than if they do it by looking at a “Wanted” pic at the Police HQ on their way to work or worse, by guessing that you have criminal intent by your behavior which can be totally innocent but look suspicious. This also reduces the need for cops etc. to intrude on people by asking for id etc. unnecessarily. Facial recognition is very accurate and will actually lead to you as a taxpayer spending less money attributed to catching criminals on the street; criminals who are probably stealing from you too. One valid concern might be an overuse of facial recognition by beat-walking cops – collection of parking fines on the spot for example might turn the public against the use of facial recognition. However, these cops are not being paid for that purpose and no rational police dept would waste their time becoming a small-fines collection agency. unfounded fears based on an irrational and unrealistic perspective is the foundation of many popular rumors and conspiracy theories.
That there is a conspiracy to invade your privacy is an interesting phobia that is prevalent among people. What these phobics get up to that is not already mostly visible through other channels is a mystery. This does not address proper espionage or secrets of ational interest; it merely address the fears of those who think they are being watched. Truth is, that this observation of your behavior happened years ago. No-one noticed – privacy invasion was inherent to the the new societal model of the millenium. CCTV just appeared and is used as a functional way to prevent us being at risk from being mugged in the street. I for one dislike the thought of being mugged more than I do that of being watched by someone I don’t know on a CCTV camera.
So hopefully this all puts the subject of Glass, Big Brother and facial recognition into a little more perspective. It does not answer all the questions for sure; an important aspect is the facial recognition and sharing of personalized data of those being recognized in real time could have major social consequences in the future. Permission to exchange data like that could well be built into the operating system and perhaps should be an integral part of all such wearable computers. What is for sure is that there will be a hundred million Glass (and other wearable) users in just a year or two and any notion that all their video could be somehow data-interrogated is more than a little far-fetched. Targeted data collection has been around for years and the tapping of major Internet fiber optic backbone by security agencies is already a known fact. As is the Prism and similar data interrogation and mining techniques. Glass is just a phone + a camera so all the data uploaded will also be passed along those same fiber optic cables through bisecting routers and fiber optic splicers into the same hands before it gets to Google anyway. If you don’t want your shit looked at – don’t go online or encrypt it, but they may have master keys for that too.
As for the processing of even a single hours’ video for 100 million users? I think that the cost:benefit ratio is far too high – exorbitantly so. Even a well-balanced economy would have a hard time paying for that. Getting Congress to fork out a $Trillion for a new Big Brother video processing server farm? That would not float in any party’s budget and would drown a party if it were an election pledge. As for the Bog Brother government or police? On the whole they don’t really have the time or the money to bother with you – if you are of interest, they probably already know about you.