[QUOTE=Kopimi;39790752]yeah i guess, but whats stopping them from doing that now? people take videos and pics of stuff constantly so they already have a huge pool of data to comb through in looking for info about people, with CCTV and smartphones and things like that. plus i guess i just can't get riled up about this sort of thing because its rather inevitable. its like, even if we acknowledge its a confirmed issue that will be taking place, what can we really do about it?[/QUOTE]
well what's stopping them now is the fact that there is too much data and too much time and effort would be expended having human beings analyze it. they already do a lot of that, to an extent, but if they develop software that can accurately and efficiently do the job for them, then it increases their ability to track citizens by a metric shit tonne.
and idk what we can do about it, but that's why you actually think about what this means for society and discuss the impacts such technology would have. maybe this sort of stuff is inevitable, maybe we can stop it, i have no idea. but i certainly think the issue is incredibly intriguing, concerning, and going to have greater and greater relevance as time goes on.
[QUOTE=gman003-main;39790684]...
3G can barely support uploading video real-time, and heavy congestion or bad reception can kill it. Not to mention how quickly it would eat up your bandwidth cap.
...[/QUOTE]
Sure, it wouldn't be plausible on 3G, but 4G (actual 4G) offers 1Gbps of bandwidth, so it becomes a non-issue with that.
Of course, providers would hate the idea of these devices streaming video 24/7, they're having problems just keeping up with people using facebook.
Edit: But I do like the idea of a company getting a subpoena to see if you ever swore.
[QUOTE=yawmwen;39790741]that's not what i'm saying. let's say, for the sake of maximum dramatic effect, the cia wants to track mr. john doe because he may be associated with a criminal organization. they take john doe's driver's license picture(along with every other picture they have), put it in the database, and up comes videos of everything that glass has caught john doe doing.
of course this is totally dramatic and not currently a real risk, but these things need to be thought about now because people are genuinely trying to work out how to feasibly do this.[/QUOTE]
I think you're overestimating how good automatic face-recognition is. It's about as good as automated translations - it works, sometimes, but it has a high miss rate *and* a high false-positive rate. Sometimes it will completely fail to recognize that a face is even in the picture; sometimes it will be unable to distinguish you from 2,000 other people with a similar face. You can fool it just by smiling hard enough, or having long hair, or wearing sunglasses. Or just by not getting a straight shot on it - most systems break down completely viewing a face in profile.
That's not something that's just going to suddenly be fixed overnight. Face recognition is one of the tougher problems, on par with machine translation, or autonomous navigation, or speech synthesis, or human-quality conversation. They can do a half-assed (or worse) job a million times faster than a human, but they can't do a good job, period.
[QUOTE=gman003-main;39790809]I think you're overestimating how good automatic face-recognition is. It's about as good as automated translations - it works, sometimes, but it has a high miss rate *and* a high false-positive rate. Sometimes it will completely fail to recognize that a face is even in the picture; sometimes it will be unable to distinguish you from 2,000 other people with a similar face. You can fool it just by smiling hard enough, or having long hair, or wearing sunglasses. Or just by not getting a straight shot on it - most systems break down completely viewing a face in profile.
That's not something that's just going to suddenly be fixed overnight. Face recognition is one of the tougher problems, on par with machine translation, or autonomous navigation, or speech synthesis, or human-quality conversation. They can do a half-assed (or worse) job a million times faster than a human, but they can't do a good job, period.[/QUOTE]
i think you are ignoring that software translators, autonomous navigation, and speech synthesis have made giant leaps forward in terms of quality just in the last decade.
yea, facial recognition is shit right now, but people are working on it and the technology will definitely improve over time.
Lets also not forget that the compression on the video and photo files coming out of Glass are going to be ridiculous. It will look alright on an iPhone screen, but anything bigger than that and a face in the background could look like anyone.
It just wont be a reliable source in any form of prosecution or employee screening.
i don't want a fucking computer in my face 24/7 so there's zero chance of me buying this
That's not even a proper article, it's a blog
This guy set out to make a point about Google Glass and its record feature, and even though there is a point to be made, he executed it horribly. His example about the bus ride is rubbish. And then he goes on to talk about your voice being recorded and tagged to you and this being somehow available to everyone through google... no.
The point to be made about Google Glass recording abilities is that, yes, you don't know if someone might be recording you, and this will bring great paranoia about people using the devices. It will most certainly be used to spy on other people and for creeps to easily record women on the streets. But regardless of this, Google Glass is a product that (the way I see it) only has ONE chance to succeed. If lots of people get it and embrace the product, it may change the way people feel about possibly being recorded completely. It may even in some ways, change society in some large-scale aspects. But if people don't embrace it, if they feel the risks far outweight the benefits (which I think it's true) then it will flop, and it will be shelved to be reopened who knows when in the future to possibly fail again, much like 3D technology or Virtual Reality headsets (both technologies which have seen recently a resurgence but will most likely fade out again).
[QUOTE=yawmwen;39790825]i think you are ignoring that software translators, autonomous navigation, and speech synthesis have made giant leaps forward in terms of quality just in the last decade.
yea, facial recognition is shit right now, but people are working on it and the technology will definitely improve over time.[/QUOTE]
It has improved, yes, but not on a significant order. It's still terrible compared to a person, and even people have room for improvement.
Honestly, I don't think it's going to become truly powerful until we get some serious breakthroughs in AI. There hasn't been a significant improvement there in years, just throwing more power behind old algorithms.
[QUOTE=Jad Hinto;39790734]I feel retarded for actually falling for that.[/QUOTE]
dont feel bad i did too
[QUOTE=Kopimi;39790658]this doesn't answer my question[/QUOTE]
Out of all the shit you've posted on the internet, your twitter, your little game development, and all your other things. You have at least once posted some information somewhere that will lead to who you are. You don't need to link your name directly to your face. All they need to do is get your real name. And where you live in. Which is apparently in Pheonix AZ, might be wrong on that. But still, its easier to find out where someone lives than you think.
[QUOTE=Jetblack357;39791049]Out of all the shit you've posted on the internet, your twitter, your little game development, and all your other things. You have at least once posted some information somewhere that will lead to who you are. You don't need to link your name directly to your face. All they need to do is get your real name. And where you live in. Which is apparently in Pheonix AZ, might be wrong on that. But still, its easier to find out where someone lives than you think.[/QUOTE]
and what if the information i've released is fallacious? what if i sarcastically said my name was dale gribble and set my facebook profile pic to justin bieber? the reason i don't take internet data mining very seriously is because it can be instantly invalidated and spoofed, it's not reliable unless it can be backed with something real and official, outside of user-generated input.
[QUOTE=Kopimi;39791073]and what if the information i've released is fallacious? what if i sarcastically said my name was dale gribble and set my facebook profile pic to justin bieber? the reason i don't take internet data mining very seriously is because it can be instantly invalidated and spoofed, it's not reliable unless it can be backed with something real and official, outside of user-generated input.[/QUOTE]
data mining isn't the end-all be-all of the search and analysis of data. data-mining simply lessens the workload by filtering text or content that you think might be relevant to whatever you are trying to figure out.
but yea, spoofing and adding irrelevant data is incredibly common and makes data-mining a lot less efficient if a group or person knows they are being tracked.
Soon enough, google will have a long database of dicks.
So it would be something like, "Oh, hey, that person has Google Glass going; quick! 'Yea, Google is sooooo awesome! I wish I could [i]be[/i] an employee! I would dedicate my whole life to such an organization!'"
Give me a fucking break, the amount of data they record for advertising along is what keeps them busy; forget any of this Glass hype. Watch and see this trend flop over.
[editline]4th March 2013[/editline]
[QUOTE=yawmwen;39791192]data mining isn't the end-all be-all of the search and analysis of data. data-mining simply lessens the workload by filtering text or content that you think might be relevant to whatever you are trying to figure out.[/QUOTE]
Through [b]a lot[/b] of data, necessary or not.
I'm guessing this is really not going to be socially acceptable to wear around other people. It's as rude as holding your phone camera up and recording everyone you're talking to. Not many people are going to be cool with you walking around interacting with them with a running camera attached to your face.
[QUOTE=tratzzz;39791193]Soon enough, google will have a long database of dicks.[/QUOTE]
Mulder, they know where your penis is, and they know what your penis is doing.
[QUOTE=tratzzz;39791193]Soon enough, google will have a long database of dicks.[/QUOTE]The government and the NWO Illuminati are going to start tracking everyone by their dicks and tits! Quick, everyone, get sex changes before they come for you!
[QUOTE=Doctor Zedacon;39791280]The government and the NWO Illuminati are going to start tracking everyone by their dicks and tits! Quick, everyone, get sex changes before they come for you![/QUOTE]
[i]"Collision detected!"[/i]
Seriously though, I can't wait for this. And I honestly like the design. I don't want it to be attached to hideous, big-framed glasses or some shitty pair of aviators or dollar store shades.
you know if there was one botnet i would not mind being a part of, it's google's.
went from cia kidnappings to nwo looking for weiners and udders in a matter of minutes
someone write a novel about this
[QUOTE=Used Car Salesman;39791255]I'm guessing this is really not going to be socially acceptable to wear around other people. It's as rude as holding your phone camera up and recording everyone you're talking to. Not many people are going to be cool with you walking around interacting with them with a running camera attached to your face.[/QUOTE]
Yeah it's basically going to be banned at all live performances, everyone's house, and most businesses
[QUOTE=IliekBoxes;39790495]This is very speculative. I don't think that all video is going to be automatically uploaded to Google servers. I think that Glass will be more like an Android device with Siri, but projected on your glasses[/QUOTE]
I can sorta answer this question
All media taken is uploaded to g+ under instant upload.
also 1080p/720p (i forget exactly)
either 83MB/minute -- 720
or 178MB/minute -- 1080
if we say its 83MB/minute
119,520MB per user per day.
if we say only 1000 people get glass in explorer edition
119,520,000MB per day or 113.983 TB.
yeah they wouldn't bother recording everything you are doing
[editline]3rd March 2013[/editline]
also: the video they showed a couple of weeks ago is [B]exactly[/B] what the glass UI looks like
Honestly this is [I]classic[/I] Google right here. Literally every product they have released has some sort of privacy-compromising feature likely to crawl up someone's paranoid ass.
The biggest issue I have with this article's theory is the manpower required to undertake such a monolithic project of tagging and sorting this archive of information.
Don't get me wrong, as we've grown into this 21st century, we've become adroit masters of digital paperwork. My part time job is data entry for researchers, and two other members of my family work as data mangers for a pharmacueutical company.
But to take the input from even a few ten thousand Glass devices, and meticulously tag, reference, archive, index and make-available certain snippets of 'interesting' or 'provocative' information is absurd. We would literally have to make the jump TO 1984, with an actual ministry of information, tagging and referencing for even the smallest areas of inflow.
God alone, imagine the nightmare of sorts a tenth of New York City's population's Glass inflow. Even with an automated service tagging faces, locations, and perhaps speech chunks, you've got little useful information there. Someone will have to manually index it for it to become meaningful. Can you imagine going over say, ten thousand auto-filed clips of "racist" content? How long would that take to tag for racism(type), racism(severity), racism(length), racism(profanity) and so on. That's even assuming you have some agreed upon index of tags. Anyone who's peered into the hivemind of an image repository site would find that taggers will spend months arguing over the most absurd linguistic technicalities of "is Onyx a rock monster, a pokemon or both? Should an Onyx tag imply them? Or does it exclude them? Is an Onyx also a snake?" Now imagine sorting that for profane or offensive material. "Is a chink really a slur? Or is it just a deragotory remark? Does it imply Chinese or general Asian? How severe is it?"
Then what about savvy "hacktivists" who wear masks, use voice spoofers, or would otherwise intentionally baffle the service? We live in an age where we fear even our most innocent phrases being turned against us. Any system so large and complex as to churn through this information would be vulnerable to sabotage of hilarious order. Imagine five hundred lulzsec legionnaires wearing masks of some influential figure, and "data bombing" his records with dozens of outrageous offensive remarks. It would be more backbreaking than protecting Wiki articles to defend against.
It's the feature no one's talking about, because it's not a feature. It's like pretending that Cellphones are actually CIA spybaubles that record our every word and allow Uncle Sam to watch us. Sure, you could make it happen. It doesn't make it a ridiculous fairytale though.
How long will it take for someone to make a John Carpenter's They Live style plugin/app?
[IMG]http://media.tumblr.com/tumblr_lq1yaehh2e1qmq2zf.jpg[/IMG]
Can't you just cover up the camera?
Stupid article, was already posted by someone else and bashed to death.
[url]http://www.thestreet.com/story/11850432/1/watch-out-for-google-glasses.html[/url]
(really, it's not a theonion clone)
[QUOTE=SGTNAPALM;39790491]How practical would it be for a company to do a comprehensive and in-depth search of an entire person over the course of [I]decades[/I]? How long does that take for a single person? How long would it take to run this check on every single one of the thousands of people applying for your company? Will this search require technology that doesn't even exist yet?[/QUOTE]
"How long would it take to run"
Google has server farms the size of a smaller country, so the answer is "in an instant"
"Will this search require technology that doesn't even exist yet"
Nope, facial recognition is dime a dozen these days and voice search can be done but it hasn't been adopted commercially (as far as I know, I could definitely be wrong here).
[editline]4th March 2013[/editline]
[QUOTE=Crazy Ivan;39793147]The biggest issue I have with this article's theory is the manpower required to undertake such a monolithic project of tagging and sorting this archive of information.
Don't get me wrong, as we've grown into this 21st century, we've become adroit masters of digital paperwork. My part time job is data entry for researchers, and two other members of my family work as data mangers for a pharmacueutical company.
But to take the input from even a few ten thousand Glass devices, and meticulously tag, reference, archive, index and make-available certain snippets of 'interesting' or 'provocative' information is absurd. We would literally have to make the jump TO 1984, with an actual ministry of information, tagging and referencing for even the smallest areas of inflow.
God alone, imagine the nightmare of sorts a tenth of New York City's population's Glass inflow. Even with an automated service tagging faces, locations, and perhaps speech chunks, you've got little useful information there. Someone will have to manually index it for it to become meaningful. Can you imagine going over say, ten thousand auto-filed clips of "racist" content? How long would that take to tag for racism(type), racism(severity), racism(length), racism(profanity) and so on. That's even assuming you have some agreed upon index of tags. Anyone who's peered into the hivemind of an image repository site would find that taggers will spend months arguing over the most absurd linguistic technicalities of "is Onyx a rock monster, a pokemon or both? Should an Onyx tag imply them? Or does it exclude them? Is an Onyx also a snake?" Now imagine sorting that for profane or offensive material. "Is a chink really a slur? Or is it just a deragotory remark? Does it imply Chinese or general Asian? How severe is it?"
Then what about savvy "hacktivists" who wear masks, use voice spoofers, or would otherwise intentionally baffle the service? We live in an age where we fear even our most innocent phrases being turned against us. Any system so large and complex as to churn through this information would be vulnerable to sabotage of hilarious order. Imagine five hundred lulzsec legionnaires wearing masks of some influential figure, and "data bombing" his records with dozens of outrageous offensive remarks. It would be more backbreaking than protecting Wiki articles to defend against.
[/QUOTE]
Any point that argues "difficulty or complexity of x" with regards to algorithms or processing power is moot. YouTube, Facebook and other services already implements filters that does advanced sound and image detection. They each handle incredible amounts of data every second. Granted, it doesn't always work and there are ways to bypass any filter. But to argue that there isn't enough "manpower" to handle the data produced by a few thousand or million Glass users is plain wrong.
Edit
To clarify, to my knowlege, the Glasses won't be streaming visual data constantly while they're turned on. Only the snippets the user chooses to upload, and perhaps the imagery that Google requires to do all the augmentation stuff.
Sorry, you need to Log In to post a reply to this thread.
