"A unique, wry and often satirical look at the Internet, the modern age and life in general" (there is way too much search-engine competition for the phrase "incoherent ramblings")

Tuesday, March 20, 2007

Human-readable captchas soon to become a reality!

New advances have been made in the technology of "captchas" which will soon render them human-readable!

If you are not familiar with what a captcha is, you are sure to have encountered them - they are the funny little graphic you are sometimes presented with when you try to log in somewhere, or post to a forum. You are asked to decipher the hieroglyphics you see on-screen and enter them in a field below, and their intention is to stop human beings from actually performing the desired action (see here for more details).

Well, it is hoped that with new breakthroughs in captchas it will soon be possible for a human being to read one of them and successfully enter the required text.

Here are some that are currently being tested and that show promise:

This one (taken from a popular forum and beginning to see common use) is beginning to show vague signs of human readability, but more work is probably required. This form of captcha would be best used to prevent entry to commoners to a forum for use by the exceptionally visually gifted.

These two examples are served up on Reddit.com and show a quite remarkable development in the concept of the human-readable captcha.

In this beta version we can see that up to 80% of the letters are human-readable, but the reading is hindered by the white letters on a white background which the scientific team behind the project is still struggling to eliminate. However, it has been shown that a 11-year old with pristine vision can in some cases correctly identify the letters. This form of captcha would therefore be useful for adult-content sites.

This captcha, used on digg.com, is probably the most impressive example of a potentially human-readable captcha. However, its readability is deceptive, using letters of different fonts and cases, thus sowing confusion in the mind of the human reader. The bleary-eyed viewer may well mistake the second letter for a capital "I", or perhaps the last letter for a small "a", forcing multiple login attempts.

While we welcome positive moves being made in the area of human-readable captchas, we would like to use the full authority and weight that this blog carries in the industry to call on science to do more!

To read the rest of this article, please type the text that you see into the field below:
(Uses Future-Phobia's proprietary HumanProof Captcha technology. All rights reserved.)