In George Orwell’s famous distopian novel Nineteen Eighty-Four, one of the most oppressive features of his totalitarian world is the ubiquitous surveillance. Every public place—and every private room as well—is always being monitored through a massive network of two-way video cameras called telescreens. Do anything that angered the tyrannical regime of Big Brother and a voice would scream at you for your insubordination.
The first time I read 1984, I had a lot of trouble with the idea of cameras everywhere. How could everyone be under surveillance? The ruling Party was only a small portion of the population. How could they watch everyone? And, of course, who watched the people who were watching the cameras? It didn’t seem possible. The math didn’t work out. The powers that be couldn’t watch everybody all the time.
The second time I read 1984 (probably around 1984, now that I think about it) I was a Computer Science student, and that gave me a different perspective. Ubiquitous surveillance could be achieved with computers.
Not the computers of the day, to be sure. Digitized sound had only just become available to consumers, and digitized video had only just been introduced to high-end production studios. Scientists were conducting experiments in the various required artificial intelligence technologies—voice recognition, natural language understanding, image analysis, facial recognition—but it was all very, very primitive. Certainly nothing useful could be done in real time.
Automated recognition of faces and voices and behavior all seemed like insurmountable problems in computer science…except that we knew it was possible because even children could do it. Two things stood in our way: Raw processing power, and algorithm design.
The first barrier was clearly going to fall. Moore’s Law—the relentless doubling of computer power every year or two—meant that sooner or later computers would have processing power equivalent to a human brain. No individual computer is as powerful as a human brain, but based on very little research, my guess is that a large modern cloud computing facility has the raw processing power and memory of several human brains. Of course, cloud data centers also use enough energy to power a small town, so we’re not quite there yet.
We’ve made some progress on the algorithms, too. Voice recognition has become a commonplace feature of some of our technology, and everyone seems to think that the NSA has computers that can pick important keywords out of telephone calls. (Although, judging by my attempts to use my iPhone’s Voice Control feature, I could hire myself out to spies and drug dealers as a low-rent version of a World War II code talker.)
Understanding natural language—spoken or written—still has a way to go. Search engines like Google can read a lot of documents, but Google only understands them at a fairly primative level. We’re still a long way from feeding the AP wire to a computer and asking it to summarize what happened today.
On the other hand, facial recognition is good enough that webcam software can find a face in the real time video stream, follow it around the frame, and replace it with a digital mask that moves its mouth and eyebrows in sync with the subject. There’s ongoing research into having computers monitor video feeds of public areas for suspicious activities.
Still, I think we’re safe from computers for now. You need humans to watch humans.
But it’s beginning to look like that may be more of a threat than I thought, and it depresses the hell out of me.
Scott Greenfield points us to the story of a blog called Just a Girl in Short Shorts Talking About Whatever. I’m not familiar with it, but apparently the author, Becky, has something of an irreverent libertarian outlook.
Unfortunately, if you click on the link above, you’ll see that Blogger (which is owned by Google) has placed a warning on the blog.
Some readers of this blog have contacted Google because they believe this blog’s content is objectionable. In general, Google does not review nor do we endorse the content of this or any blog. For more information about our content policies, please visit the Blogger Terms of Service
[And if you click “I understand and I wish to continue” you’ll hit a page that plays music, so be careful if you’re at work.]
The content warning is devastating because it apparently interferes with the search engine indexing of her site—something that a Google-owned operation like Blogger should understand. Becky says she’s vanishing fast from Google’s search engine. (On the other hand, Google is still making money off of AdSense advertising on her site.)
Becky attributes this to “some rather low-level Google employee” responding to complaints. I have my doubts. Google likes to automate as much as possible, so I suspect the flagging of the content as objectionable is done entirely by algorithm. Here’s what Google has to say on one of their help pages:
When someone visiting a blog clicks the Flag button in the Blogger Navbar, it means that person believes the content of the blog may be potentially offensive or illegal. We track the number of times a blog has been flagged as objectionable and use this information to determine what action is needed.
I don’t see anything about a human review. Google would prefer it that way so they can save labor and claim to be unbiased. In this way, labeling the content as “objectionable” is a purely objective designation: People have objected to it.
(If you’re thinking of leaving a comment pointing out that libertarian principles allow a non-governmental entity like Google to censor whatever they want, please know that—as is typical of libertarian debate—me and Scott Greenfield and the guys from Popehat have been spraying each other with friendly fire all day over this issue in Scott’s comment section.)
In any case, Becky also touches on a far more disturbing issue:
I am pretty certain that any readers, who took the time to bitch and moan to Google, were much more concerned about my politics, and ideological bent, than anything that actually offended their prurient sensibilities–this has historically been the case with this sort of thing on this blog.
I’ve been around the web long enough to hear a certain ring of truth in what she’s claiming. Basically, her readers didn’t like what she was saying and—rather than just reading something else—they decided to manipulate Blogger into silencing her.
Sigh. When there are cameras on every street, in every business, and in every room of our private homes, Big Brother won’t need a hundred warehouses full of computers to keep an eye on all of us and punish us when we stray from the party line.
The computers might help, but we’ll do it to each other.
Leave a Reply