reCAPTCHA is a free service that adds a magnanimous twist to the whole, “prove to me you’re human” paradigm of anti-spam technology. It was a novel idea back in the day, I’m sure, almost taking a page from the Turing Test. I can remember the first time I used one. I thought it was a great idea—very neat way of identifying humans from spambots. Then, new ones came out. More advanced ones. Some more difficult to use than others. Some downright annoying. Some did a better job than others. And for a time, things were good.

I have alway wondered whether some of the more difficult ones were made that way because spammers figured out ways to circumvent them, or if it was just overzealousness on the programmers part—trying to make an anti-spam system that could never be beat. In any case, it looks like reCAPTCHA has really gone into overdrive.

I admit that I installed the service on my own site (and for most of my clients) primarily because it helps digitize books. I think that’s a terrific idea. I can’t tell you the wealth of information that exists in the scientific community, but most of which is still relegated to physical journal articles. I would much rather have a hardrive full of PDFs, than a closet full of boxes. The trees we save not withstanding, it also helps make knowledge dissemination much easier and fluid. A win-win in my book. I am all for the spread of knowledge, and if I can protect myself and my clients from spam and promote the distribution of information, then just show me where to sign up.

But it looks like we hit a chink in the chain. In all honesty, reCAPTCHA “challenges” have never been the easiest. Every now and again, I’d get one that was perhaps a little challenging; something that didn’t just pop out at you. Then they added “blotches” behind the words, which has inevitably made their system far less attractive. Here you can see just how difficult some can be. Now I didn’t cherry pick these in an entire afternoon. This list was taken from around 15-20 reCAPTCHA challenges. That means a considerably high number of challenges are far too difficult, while some are just downright impossible (the last one eludes me completely, I want to say the first sequence is “im”?).

For a native English speaker, most of the challenges can probably be deciphered with minimal effort, despite the new blotches. Being familiar with a language affords you the advantage of knowing what letters you are likely to expect. In fact, research into cognition and perception have illuminated researchers on how the human brain processes letters. Wchih is why rideang steneces lkie this, aern’t all taht hrad. If we see the letters B-E-A… most likely we are going to presuppose that either a T or an N or an R may likely follow. If we see something that looks like a vowel, or an H, we would be required to expend a little bit more cognitive power in deciphering the word. This is also the case when reading full sentences. We come to expect certain words, and if they don’t follow (through typos or just bad writing), it becomes difficult to read. It’s like having to stop at every house to see if it’s the correct address. Just not productive.

However, the world is a large one. The internet has been breaking down the walls between countries for some time now, and while I’m sure the vast majority of people that frequent my sites are indeed fluent in English, they may not all be. Moreover, English may not be their primary language. If that’s the case, then they may not be as comfortable or as adept at decipher English text. Sure reCAPTCHA provides access to other languages, but as far as I can tell, it has no “multicultural” setting. It’s either English or German, but not both.

If it were a simple system that identified spambots, I would have removed it, but I am still wanting to digitize those books. Yet, I also care about my visitors and those of my client’s. And what’s more, I am also a usability freak. I don’t feel the current reCAPTCHA interface provides a clear enough “refresh” button to create new challenges. Moreover, I don’t think that’s an adequate solution. Forcing visitors to continually cycle through challenges until they find one they can accurately transcode doesn’t jive with productivity. I think it also asks too much of people. Let’s face it, people want to send you an email, not decipher a Sherlock Holmes mystery. I would like to make that process as easy and as noninvasive as possible. In an ideal world, they should be able to click a button and fire that email; all this spam protection should be well behind the scenes.

With the advent of social network and the blow up of blogs, the nuisance these kinds of devices elicit has increased exponentially. Most blogs or forms require some form of human validation. Well, I for one, am getting really sick of these time consuming processes. I know I’m human. I feel like I’m in some awful universe that’s overrun with militant robots, forcing the remaining humans to constantly present their IDs wherever they go. If that’s the case, then it seems to me we have already lost the war on spam.

So let’s get back to the drawing board. Sure these types of anti-spam measures worked when the internet was a static place to visit. But it has matured. It is full of dynamic content. Moreover, it has become a real communication platform. Information is exchanged readily and drives commerce to real relationships. Let’s bring down spam, but in a sensible manner that still respects the human behind that keyboard. After all, time is exceedingly precious.