Online: | |
Visits: | |
Stories: |
This is one form of electronic surveillance that is entirely worth it.
By now we all know that horrible test common to sites that want to weed out spammers and robots. You’re presented with a block of garbled, nearly-indecipherable text, and asked to type it back to prove you are human.
Well it turns out robots are pretty good at pretending to be human, and getting better all the time.
Google, which is one of the biggest providers of CAPTCHAs (the garbled text tests), has a new system that pre-screens and approves most humans. According to The Verge, the company looks “at factors like IP addresses and time spent on page” to figure out if you are real.
And instead of presenting the standard text challenge to those who don’t make it through, Google is trying new ideas, like a grid of similar images based on its image search product.
One would think the company that reads our email in order to figure out what kind of underwear to sell us would be better at knowing if we’re human. It turns out, they are.
—Posted by Peter Z. Scheer
Related Entries