(Disclaimer: I work at Google, but not on ReCaptcha.)
The point of this change is to make things easier on 90% of humans -- the ones who have JavaScript and third-party cookies enabled now get to tick a checkbox and be on their merry way, instead of doing a useless captcha when we knew they were already humans. Recall that when ReCaptcha initially came out, the argument was "humans are wasting all of this time, let's turn it into useful work to digitize books".
If book-based or street view-based captchas go away, I suspect it will be because bots/spammers got better at solving them than humans, not because Google thinks that the machine learning spam detection approach is fail-proof.
Recall that "reading" captchas already pose an insurmountable barrier to users with conditions such as illiteracy, low vision, no vision, and dyslexia. To accommodate these users, audio captchas are also provided, but a 2011 paper suggests that audio captchas are either easy to defeat programmatically or are difficult for users themselves to understand: https://cdn.elie.net/publications/decaptcha-breaking-75-perc...
I am visually impaired and can attest to both visual captchas being a pain and audio captchas being hard to understand. this change is nothing but an improvement as far as accessibility and usability goes. This is only a plus for people who implement these, as I have actually left sites that had insurmountable captchas for me.
"WebVisum is a unique browser add on which greatly enhances web accessibility and empowers the blind and visually impaired community by putting the control in your hands!"
"Automated and instant CAPTCHA image solving, sign up to web sites and make forum posts and blog comments without asking for help!"
I was curious about the CAPTCHA solving, too, so I tested WebVivum out on ~8 reCAPTCHAs.[1] It solved all except 2 of them, taking 20-60 seconds each time. In 2 cases it reported failing to solve the CAPTCHA, but it never gave an incorrect result. That is, whenever it gave a solution the solution was correct (in my brief test).
So, while it's some way off their claim of "instant" CAPTCHA solving, this is definitely a very useful addon, especially for those people who cannot solve CAPTCHAs at all. Thank you for pointing it out.
There are ways to solve captchas somewhat reliably programatically. I suspect this plugin only works on certain computer generated captchas, not the street sign ones.
Very helpful, thank you! I have a difficult OCR problem to solve, rather than identity. Interesting to see that the market price for "being human" is $0.00139.
The fact that this works shows that distorted-text captchas are no longer effective.
From the Google's blog post:
> our research recently showed that today’s Artificial Intelligence technology can solve even the most difficult variant of distorted text at 99.8% accuracy
If book-based or street view-based captchas go away, I suspect it will be because bots/spammers got better at solving them than humans
But, wait. Isn't that what we want? It seems like bots and spammers have a relatively small cost to a company like google, while digitizing books and house numbers is relatively valuable. I don't have numbers for a detailed cost-benefit analysis, but if bots get good enough to do time consuming work accurately, that's a win right?
That's like flying because you like airline food. No one flies if they don't have a destination. No one will put a captcha on their site if it doesn't tell computers and humans apart; that's its primary job.
The point of this change is to make things easier on 90% of humans -- the ones who have JavaScript and third-party cookies enabled now get to tick a checkbox and be on their merry way, instead of doing a useless captcha when we knew they were already humans. Recall that when ReCaptcha initially came out, the argument was "humans are wasting all of this time, let's turn it into useful work to digitize books".
If book-based or street view-based captchas go away, I suspect it will be because bots/spammers got better at solving them than humans, not because Google thinks that the machine learning spam detection approach is fail-proof.
Recall that "reading" captchas already pose an insurmountable barrier to users with conditions such as illiteracy, low vision, no vision, and dyslexia. To accommodate these users, audio captchas are also provided, but a 2011 paper suggests that audio captchas are either easy to defeat programmatically or are difficult for users themselves to understand: https://cdn.elie.net/publications/decaptcha-breaking-75-perc...