End of the Road for CAPTCHAs

Signing up to a forum the other day, I couldn’t help noticing the increasingly pervasive presence of anti-spam and anti-bot modules built in to online registration forms.

The problem with CAPTCHAs in my opinion is simply that they are stressful to use for a modern end-user. Research (c.f. some New Scientist article that I read) indicates online users have over several years become far more ‘aggressive’ in that website visits are increasingly about simply extracting information and leaving: people are becoming more ‘hit-and-run’ as opposed to ‘come join our wonderful community at Experts Exchange’! This speed and efficiency (which I think is largely attributable to Google) of surfing however goes completely out the window when it comes to CAPTCHAs: rather than simply clicking ‘OK’, users are forced to squint at a series of curly, obfuscated, low-contrast, specked, half-obscured characters, often only to be told they’ve typed the wrong characters and forced to repeat the whole unpleasant experience. CAPTCHAs break the line of thought and direction of a user, serving merely to annoy him/her; most people I know only ever think of CAPTCHAs as infuriating and time-consuming wastes of space. They’re sometimes even vaguely funny!

There’s actually an argument that the whole concept of forcing users to read half-illegible characters is doomed anyway. There are very, very good programmers and hackers out there who are working towards better text-recognition software. ABBYY Finereader is one of the leading Optical Character Recognition (OCR) programs out on the market, and Evernote wasn’t deceived even by my handwriting – technology has surpassed most of the humans I know! I am quite convinced that there will come a day on which computers and bots will eventually become better than humans at recognising characters and completing optical identification tests. I’m pretty sure CAPTCHAs seem to get harder and harder to the (ridiculous) point at which it takes several attempts to get one right, for example the ‘cat’ CAPTCHA at Rapidshare:

I became an expert at doing these and I assure you the answer is KOI8

I became an expert at doing these and I assure you the answer is 'KOI8'

and this one which I think is a fake but is far too good not to include:

So CAPTCHAs are ultimately heavily limited in their applications, especially considering computers are growing in power at an exponential rate (Moore’s Law). Alternatives often offered to CAPTCHAs involve having the characters read out through the speakers. Unfortunately I think speech-to-text software is also at a highly advanced stage (so long as Vista’s attempt at it is ignored).

So, have people had any other workable anti-bot systems? I certainly think so. In fact the example I gave near the top was a good example of such an alternative: a turing test. Programs which have the ability to ‘parse’ English and make decent ‘sense’ out of it are still things of science fiction. Admittedly there are some very limited versions of artificial intelligence, however those tend to cost rather a lot more than the average spammer can afford, and your average bot is still a long way from answering questions like “is ice hot or cold?”

BT: is ice hot or cold?
SmarterChild: People always ask me if there’s a smarter bot than SmarterChild. Well, now there is! Add wilma@live.com to your Contact List to see what I mean!

During the course of our conversations you may see sponsored messages. These messages have the word “sponsorship” at the end. Sponsors help me remain free of cost to you. To learn more about sponsored messages, click here —> http://web.smarterchild.com/ct?xCkO/0

BT: is ice hot or cold?
SmarterChild: I’m sorry, I don’t know what you’re saying.

*Adds wilma@live.com as a contact*

BT: is ice hot or cold?
Wilma: I am sorry, I didn’t get that.

I’m just getting started, so I can’t answer everything yet. I’m getting smarter all the time!

Would you like to try a search for “is ice hot or cold”?
BT: you know what, never mind…

I have to say though, my favourite proposed alternative is this one:

Perhaps the people who designed it were attempting to guarantee their site an intelligent user-base!

So anyways, I think the CAPTCHA is near the end of its life. When confronted with a picture of what appears to be a random doodle and expected to use their artistic licence to deduce some vague form of meaning from it, I’m pretty sure many people in a hurry would be annoyed, end up wasting time getting the CAPTCHA wrong multiple times and end up losing track what he/she was supposed to be doing when signing up/posting in the first place. I hope the future holds something easier to use for users of t’internet, much as I love the idea of getting users to solve Maths questions before allowing them to post comments. Perhaps ultimately Man will lose his battle against the machine and whatever anti-bot system ends up being put into place will be solved faster and with greater accuracy than humans! For now though, most machines powerful enough to perform complex human intellectual feats in any reasonable amount of time occupy several hundred square metres of floorspace in labs at NASA and IBM. I sure hope NASA aren’t secretly spammers…

Interestingly, a productive use for CAPTCHAs has actually been found. For those who still haven’t come across them, reCAPTCHA systems have been deployed all over the web. As stated on the reCAPTCHA website:

reCAPTCHA improves the process of digitizing books by sending words that cannot be read by computers to the Web in the form of CAPTCHAs for humans to decipher. More specifically, each word that cannot be read correctly by OCR is placed on an image and used as a CAPTCHA. This is possible because most OCR programs alert you when a word cannot be read correctly.

CAPTCHAs were useful once, back in the year 2000 when computers ran on processors just about capable 7 flops and could barely run fortran compilers (yes I am being ironic). In fact, they were quite a good, and rather an ingenious, solution to the problem, as before they began to get ridiculous they were convenient to completely and only ever took about 2 seconds, and computers were more or less baffled by even slightly complicated/curly/irregularly typed/written text versions. But in a modern world I think the whole process of exacerbating myopia is bloated, unnecessary and pointless. But then maybe that’s just me!



Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: