Like most people in the tech industry, I’ve always liked Google. From the moment I typed my first search query into the no-nonsense search page, they’ve been able to do no wrong: I like everything they do, how they do it, and why they do it. Their services are so reliable and useful it’s very difficult not to think of as being an integral and always-available part of the infrastructure of modern life, like electricity or banks. I do, however, read quite a lot of technology news, and various bloggers and pundits always seem to eventually get to the point where relying on Google burns them – sometimes pretty badly, in the major livelihood compromising sense – and they have to scramble to reorganise.
I got my turn a couple of days ago. Over the last few years, I’ve prepared maybe a dozen or so short demo videos showing how to use my mobile apps, for tasks such as drawing structures, which I uploaded to YouTube and linked to from my company website. Also a couple of recorded presentations (voice + slides) and webinars, too.
Some time ago, I prepared a novelty demo, which showed how to draw the chemical structure of viagra, using one of my mobile apps. It was intended to show that the time required to use my advanced sketching user interface for touch-based devices is quite competitive with the equivalent process using a traditional desktop-based interface. Quite harmless, you might think, until I got this in my inbox:
Doesn’t sound all that unreasonable: maybe a spambot analysed the content and found the keywords to be quite compatible with content that’s not supposed to be there. Something in the wording of the message suggests that it percolated up and got looked at by a person, and that there would be opportunities to dispute this misclassification. Perhaps it is related to a similar observation that WikiPedia considers porn stars to be far more notable than scientists: is this a scientific literacy issue? Or is it something even less palatable?
First of all, I’m pretty sure no human was involved: at least, not a person who passes the Turing test. Secondly, the email was actually highly misleading: what had actually happened was that this single automated flagging shut down my entire account, immediately, with no warning, no way to do anything about it. The methods for appeal didn’t work, because my account was terminated: I need to be able to login in order to do that. All of the videos that I uploaded are now deleted, not just the misclassified one: they have been removed from the internet; it is as if they never were. My demos and presentations are now unvideos.
Needless to say I clicked on all the appeal buttons that I could find, and wrote them a message explaining that if somebody with a basic knowledge of science were to take a quick look at the offending video, it would be immediately obvious that there’s really no problem, and that the accusation, trial & sentencing should all be repealed.
Franz Kafka’s classic dystopian novels described what happens when an uncaring bureaucracy is activated for the arbitrary persecution of individuals without a legitimate reason. A hundred years later, we get to experience a similar process, except rather than cold heartless government officials following the procedures of the law, now it’s computerised algorithms looking for patterns, and executing the entire process without any moderation.
In my case, this is not exactly the end of the world: a few clickthroughs from YouTube isn’t going to determine the success or failure of my business, and I promptly reuploaded most of the demos to Vimeo, and updated the links to them. It’s annoying that the indexing and crosslinking has vanished, but I’ll get over it. There are a lot of people out there who have experienced the same kind of trial, some of whom have suffered significant loss of income.
The moral of the story is: don’t get too dependent on any company, or take their good behaviour for granted. When it comes down to it, no company can not be evil forever.