4 ms·
Ask HN: How close are we to a point when Google can ruin your life?
Hardly a day goes by anymore without Google committing some new outrage against personal privacy...even that of non-users. The London Review of Books - http://www.lrb.co.uk/v33/n19/daniel-soar/it-knows - has one of the better roundups of the current state of Google's surveillance against its users.
I'm not here to rehash the online privacy debate, but to ask a new question. One of the things implied by the article is that given the voiceprints and images Google already has on file, it's relatively trivial to write an app that can identify anyone within a minute or two; the only thing actually preventing it from happening right now is that the API's either don't exist, or aren't available to the public.
Another thing implied by the article is that Google itself is attaining a certain level of sentience, or at least an ability to understand the contexts and meanings of what it reads. Google is improving itself constantly by absorbing and collating the information it receives through its billions of sensory organs. Presumably it's doing the vas majority of this without human intervention.
And then there's this in New Scientist - http://www.newscientist.com/blogs/onepercent/2011/08/facial-recognition-identifies.html - which basically says that Carnegie Mellon has an experimental app that runs the same facial recognition and, once it's identified you, tries to find your place and date of birth in public records; which, once found and taking it several steps further, allows the app (or another app) to make a reasonable guess as to your social security number.
Now. Google asked a friend of mine for her date of birth today, before letting her get back into the Gmail account she had for the last seven years. (It's hard to imagine a 6-year-old opening a Gmail account in 2003, but apparently they were worried she was under 13). What's interesting is that it didn't ask for my DOB. So I went to Google's Privacy Pages, which purport to list the information Google has on file about you. Nowhere in those pages was my date of birth. YouTube has me at 31 years old, but doesn't show my DOB itself. Which is interesting, because they must know what it is.
Nor is there anywhere in Google's privacy pages where it shows the biometrics of my facial features, or my voiceprint, although presumably those are also on file. In fact, there's very little real information shown there.
So, all that being the background, here's my question: What if Google -- the sentient thing out there, not the puny corporate hacks who are nominally in charge of it -- decided to ruin your life? I don't mean by de-ranking your website; I mean by really, seriously screwing with you? Could it forge your identity and rack up credit card debt, making it seem like the perpetrator was someone in Nigeria? Could it fake a photo of you cheating on your wife? Could it take over your bank accounts, guess your other passwords, or subtly alter your outgoing gmail messages to make them offensive yet plausible? How far are we from the point where it could do those things on its own?
And lastly, how close are we to the point where writing "fuck google" on the google forums will cause it to lose its temper?
- anigbrowl 15y agoHardly a day goes by anymore without Google committing some new outrage I suggest you try this again without the axe-grinding editorial slant. By mentioning only a single company in a context which could be applied to several large firms, I can't help forming the strong impression that you work for a competitor.
- noduerme 15y agoplus one I started writing this originally including Facebook, which to my mind is the other really egregious violator of privacy, and probably worse than Google. However, I'm not a Facebook user. I have been in the past, and they still collect data on me, but I realized Facebook doesn't have nearly the complete picture of my life that Google does. Nor do they have my mail, my spreadsheets, my contacts. Most of all, Facebook is not in and of itself, a machine potentially capable of abstract thought. That was the heart of my question, and I'm not sure if you made it past the first sentence and read that far. I wasn't asking which corporation would screw you over with the privacy you gave up -- I'm not interested in grinding a personal axe or rehashing the online privacy debate. My question, as phrased in the subject header and as borne out through my post, was whether Google AS A MACHINE was capable of ruining your life, and if not, then when it would be and how it would proceed to do so.
- anigbrowl 15y agoI did read the whole thing, and I did find that question interesting. But honestly, I felt it came off as a bit pissy, and also underestimated the potential for other software behemoths to cross some AI boundary. I suggest rereading (or reading) Neuromancer and thinking about the differing approaches of the two AIs referred to in the book. One is very factual and logical and can manipulate some things directly, another is weak in that area but understands people well enough to engineer coincidental-seeming meetings and interactions that it anticipates will be helpful at some later date.
- icebraining 15y agoA 'sentient' Google? Frankly, that sounds paranoid. Google has algorithms, developed by humans and which don't change unless someone edits them. The idea that it can somehow get 'pissed' and start purposefully trying to ruin your life is ridiculous. Google 'grows' in data and it certainly improves certain AI-like mechanisms, such as spam detection, face and voice recognition, etc. Programs don't magically become sentient from that.