The New York Times had a great article concerning the bane of “Completely Automated Public Turing Test to Tell Computers and Humans Apart” — “captcha” for short — on our virtual lives.
If you don’t recognize the word “captcha” you will certainly recognize their semiotic power over your online life:
I hate captchas. I can’t read them. I hate being punished for being human.
Some blogs use captchas to block unwanted comments but, as you now know, the dedicated Spammer can easily get around any captcha locks.
The “audio” captchas are the worst. Have you heard them? Find a difficult captcha — Google
has a few for identification verification for some services — and
click on the “Handicapped” icon next to the captcha.
One would think an audio captcha would be a spoken version of the
semiotic version but that is not so… audio captchas are a beast of
their own and consist of softly spoken numbers voiced over white noise
and a cacophony of other random numbers spoken in the background.
Your job is to figure out which numbers to listen to and then them type
into the captcha decryption box.
The experience is enough to make you
recognize how it feels to go insane.
The good news in the Times article is that “relational” captchas are on
their way. You’ll be presented a series of images — tree, sock, hat,
orange — and you would have to click on the image “you wear on your
Since there are millions of images one could use for the captchas, the
theory goes, it will be harder for the Spammers to fake human sorting values
and quantifiable judgments.
I suppose it’s better to click on the right image than to try to type
unreadable characters or enter inaudible numbers — but I wonder if any
captcha technology is really worthwhile.
Do captchas punish Spammers or only us regular folk?
How many captchas do you have to enter in a day?
Does captcha technology ever push you away from a site or a service
because you don’t want to have to deal with repeated failures in trying
to decode the captcha lock?
I loathe them – I encounter them mostly on the new blogger/google blogs – long incomprehensible strings in excruciatingly complicated fonts. I give it three goes then give up.
They seem to be getting longer and longer – why can’t they just work with Akismet like WordPress do?
Yes they do put me off commenting at times – and are guaranteed to wind me up!
I loathe them as well. I prefer to have people register once to be able to place comments than to have them deal with a captcha for every single comment they want to make.
Akismet is owned by Automattic — so I’m sure Google/Blogger and other similar sites don’t want to admit WordPress.com does Spam protection better than what they offer.
Some WordPress.com users have complained about a LACK of a captcha system to block Spammers and the WP.com gurus have flat-out said, “captchas don’t work.”
They do seem to get longer and more skewed. It’s impossible to deal with if you’re in any sort of hurry.
I think trying three times is very generous. I wouldn’t try more than once. 😀
Spammers suck…but captchas blow!
I once lost some really sweet seats on Ticketmaster when I could not decode their stupid captcha. I also tried to listen to the audio captcha and, you are right, that was even worse! By the time I figured out what the heck it was saying, my session had timed out and I lost my tickets. Grrrrrrrrrrrrrrrrrrr!
I thought I was the only one who couldn’t type the right response to captchas, especially if the are case sensitive. I’ve had problems correctly answering captcha using Google’s Blogger, Yahoo, and MySpace — those are probably the big three of the internet. Myspace’s captchas are probably the easiest the read with Yahoo’s being the hardest.
I can’t imagine having to listen to a captcha — I’ve never done it and I doubt that I’ll try it.
It find it is hilarious that machines are able to figure out the answers to captchas when humans often have a hard time figuring out the right answer.
Yeah, they do BLOW! So why are they popular? Why are we stuck with them when it seems it has been proven the current system does not work?
I had the same sort of experience trying to buy tickets online for The Vagina Monologues — my golly that stupid timer! They give you the worst seats first and it takes three or four “dig” requests to find better seats and by that time your time is running out!
Then… they give you the captcha to decode. Grah!
I’m with you. If I see a captcha I turn around and run the other way if I can. They are extremely hard to figure out — like “i” and “1” and “7” — I sometimes give up trying altogether.
You should try an audio captcha! I had no idea what I was getting into when I tried that method. There are no instructions on what to listen for and all your hear are a bunch of numbers being spoken and you’re trying to write them all down and then VERY FAINTLY one number stands out above the din of the others and ONLY THOSE NUMBERS are what you’re supposed to key into the decode box. Who knew? It is pure insanity — especially since audio captchas are supposed to be “handicapped accessible”… but that provides no relief for the Deaf or Hard-of-Hearing.
I agree it’s funny machines do the decoding better than we do.
I read somewhere that solving a simple math problem would work even better than the current image captchas– but I’m not sure how — because machines do numbers better than people!
I wonder what the solution is? Captcas don’t work to keep spammers out, but make it difficult for many humans.
I think the Askimet system is probably the best compromise — even then, that system keeps legitimate humans from participating at times.
I agree with Chris that Yahoo’s captchas are the hardest to read. I tried unsuccessfully to read one yesterday and finally gave up. Little things can be so very frustrating!
Hi Chris —
Yeah! Captchas only keep out the good, but easily frustrated, humans. The Haters and the Spam bots always find their way inside. 😀
I think the current captcha system is much like those “Home Alarm” signs you see staked on front lawns and stuck in windows — 90% of them are just for show and there’s no active alarm system in place and the captchas serve the same purpose for the uneducated: “See? We have these captcha locks on our front door to protect you from all the bad people!” 😀
Akismet is pretty good. I much prefer registration and verified IDs. Our hater/nuisance posts have dropped off a lot.
I also think the “which image doesn’t belong” is also quite interesting because there’s no way for a machine to make that sort of judgment quick enough to be effective on a big Spam basis.
Ouch! What were you trying to do on Yahoo! that forced you to give up?
We could have a SAT-type question:
Please read the following paragraph and pick out the answer from those given that best answers the question. If you find that there are two or more answers that could answer the question, please select the option to select multiple answers. You have 10 seconds to answer the following question.
Henry just arrived on Flight 249 at Chicago’s Midway Airport. Paul is at O’Hare Airport and is scheduled to depart on Flight 976. Henry wants to take the Chicago Transit Authority’s train from Midway to O’Hare Airport. Paul wants to purchase some alcohol from the duty free store. How many pennies does Bob have in his pocket when he shows up at the security checkpoint at JFK Airport in New York?
A. 2 P.M.
B. Blue Line
C. Orange Alert
D. 29 cents
E. None of the above
Harr! I love that! I never tested well, so that solution will completely lock me out of anything on the net. 😀
I thought at the end you were building up to ask this old chestnut: “What is the name of the bus driver?” 😆
I should have asked something like that! I’m sure if there was some sort of comprehension test designed to filter people from computers, we’d all be scratching our heads trying to figure out what was being asked.
I took a look at the sample SAT question of the day to see what kinds of questions they were asking. The scary thing is that they aren’t that much different. 🙂
Questions like that are meant to fool. I know there’s a way to “figure out” what’s being asked but the tricks are rarely obvious or opportune and the creative, insightful answer is always wrong in favor of the mundane!
I was trying to logon to Yahoo! Games to check for a confirmation email on a video game I ordered for my nephew. I had just set up the account and misplaced my ID.
I was in a hurry and it wasn’t worth the frustration to try to figure it out, so I gave up.
Found my ID and password today so all is well. 😀
I think your experience is a dangerous lesson for places like Yahoo! that force good people to perform by jumping through silly hoops.
Now that’s funny!
What a great find!
How did you find it?
I plugged in “captcha reading service” into Google to see what might pop up. I figured someone probably hires people — via a make millions working at home with your computer offer — to type in various captcha, so I wanted to see if that type of service would pop up.
Now, if you outsource captcha reading overseas where people get paid a few pennies per hour, the cost per captcha read could be decreased ….
Don’t reveal all your secrets! 😉
I get all my images from Google Maps on the Street Level.
I also go to museums and then peer in their windows via Google Earth and screenshot all the great paintings in antiquity! 😀
Street level Google maps are pretty cool!
I just heard a report that cell phone companies might help with traffic reports by telling where all of their cell phone signals are congregating by using the subscribers’ phones GPS.
Hi Chris —
Yes, Google Street Maps are definitely keen. I think Yahoo! and Microsoft beat them to the feature but Google does it better than both of them.
GPS in phones is killer. It would be awfully easy to exploit, though. So far people have the option to turn it off if they don’t want to be followed.
I read about a captcha system that is pretty smart, but isn’t impossible for humans to figure out. They show pictures of animals as the identification tool.
Another cool thing with MS’ Asirra animal identification system is that you can adopt the animal, if you want! They get the pictures from Petfinder.com.
Yes, that’s the system mentioned in the NY Times article! I’m so glad you found it! There are millions of images of pets so it would be almost impossible for Spam network to index and identify them all.
Oh, and I forgot to mention Chris, that I took the sample test in your link and while it was easy, it still took too long. I’d prefer to pick two cats out of six than six out of 10.
I was having fun with also. I was selecting dogs, instead of the cats. The computer accused me of being a bot!
Chris! That’s funny!
I was stumped on one image — it was a shot of a dog’s rear end. I wanted to do well so I really had to look at it since the only thing I could see was the fur and the hind legs. I decided the hair was too curly to be a cat and I was right. But I spent 20 seconds deciding! Too long!
To commemorate the 17-year return of cicadas in Illinois, they could have various pictures of the insects to select. Later in the summer, they could have mosquito pictures. 🙂
Love it, Chris!
I think discerning between images are a great way to fight Spam and because images — SEMIOTICS — are the universal human language, the system is fair to everyone.
I wonder if we could create our own — Urban Semiotic Captcha Gotcha System — where we’d use our blog Author Avatars and images of cats and dogs and then ask commenters to “Pick the animals.”
How many perfect human scores would we have?
We probably could do it — someone should grab Microsoft’s API. I’m sure there’s a provision to point it to another source for additional images.
But if someone selected the “Chris Hedges” Avatar as an animal — would that be a right or a wrong answer? 😀
Lol … we’d have to give them credit for both 🙂
Heh! I think you’re right, Chris!
what i hate in this system is that most of the time they make me confused about what is an upper case letter and what is not
They are confusing to figure out just as you say!