I remember the first time I saw a computer with a microphone attached to the monitor — it was 1991 or 1992, and it was a brand-new Mac on the department secretary’s desk at a major university. I asked her what she used it for, and she said, “I don’t know how to use it yet — it came with the new computer.” I was immediately intrigued and a bit alarmed — this was in the days when the internet was ubiquitous on university campuses, but virtually unheard-of by the general public. And computer security was very rudimentary. My first thought — literally, my first thought — was that somebody, somehow, could probably use that microphone to eavesdrop on conversations in the office.
And now, 15 years later, Google has plans to do just that.
The Register is reporting that Google is going to deploy software — “sooner rather than later” — to listen in on users, analyze the sounds in their environment, and serve up appropriate advertisements:
The idea is to use the existing PC microphone to listen to whatever is heard in the background, be it music, your phone going off or the TV turned down. The PC then identifies it, using fingerprinting, and then shows you relevant content, whether that’s adverts or search results, or a chat room on the subject.
And, of course, we wouldnâ€™t put it past Google to store that information away, along with the search terms it keeps that you’ve used, and the web pages you have visited, to help it create a personalised profile that feeds you just the right kind of adverts/content. And given that it is trying to develop alternative approaches to TV advertising, it could go the extra step and help send “content relevant” advertising to your TV as well.
Now a lot of people find using personal information to deliver ads offensive to their privacy. I am not really scared of ads, but I’m scared of other uses the same technology could be applied to. As The Register points out:
Pretty soon the security industry is going to find a way to hijack the Google feed and use it for full on espionage.
Google says that its fingerprinting technology makes it impossible for the company (or anyone else) to eavesdrop on other sounds in the room [besides TV], such as personal conversations, because the conversion to a fingerprint is made on the PC, and a fingerprint can’t be reversed, as it’s only an identity.
This is complete baloney. Sure, maybe the currently-proposed version just listens for TV and just sends information about what show is on, but that doesn’t mean someone else — at Google or otherwise — couldn’t use the technique to capture the actual audio content, or even an automated transcript of personal conversations. With sufficient data, they could even use audio “fingerprinting” to determine who’s talking — even if they aren’t using the computer.
Moral of the story: Unplug your computer’s microphone.
Plug it in only when you need it.
That may be never. Personally, I’ve been waiting 15 years for an actually useful purpose for the PC microphone. The potential seems endless — Internet telephony, voice chat — even encrypted voice chat, voice recognition instead of typing, voice annotations on documents, etc. But it never seems to pan out. In all that time, I’ve known only one person who ever had a use for that, and it was voice recognition instead of typing — because she had a wrist-pain problem. The system was OK, but it wasn’t good enough for her to abandon typing when her wrist pain wasn’t flaring up, and it wasn’t good enough to capture significant market share among people without severe wrist pain.
I don’t know, maybe the rest of you all use voice IM, and I’m just behind the times — but if you do that, unplug your microphone when you’re not using it.
And be careful.
Big Sibling is listening to you.