Are You A Robot?

[Eric]

I was just presented with one of those ‘Are You A Robot?’ Challenges, during a login. I was instructed to click on all the boxes containing part of a motorcycle, only the picture didn’t contain a motorcycle, it contained a ******* scooter!

So, my bank just forced me to identify a scooter as a motorcycle, in order to access my bank account. I will not forget this needless and humiliating indignity, you bastards!

[Duluth]

[pittpa]

Triggered ?


Sent from my iPhone using Tapatalk

[Eric]

7 minutes ago, pittpa said:

Triggered ?


Sent from my iPhone using Tapatalk

Damn right!

[Pima Pants]

Are we not men? We are DEVO.

[Maser]

I definitely prefer those over those obnoxious ones where you had to copy the letters and numbers shown to you, but they were so deformed you couldn't even tell what they were. 

[Duluth]

5 minutes ago, Maser said:

I definitely prefer those over those obnoxious ones where you had to copy the letters and numbers shown to you, but they were so deformed you couldn't even tell what they were. 

[Eric]

1 minute ago, Maser said:

I definitely prefer those over those obnoxious ones where you had to copy the letters and numbers shown to you, but they were so deformed you couldn't even tell what they were. 

One of the technologies used for this purpose was an interesting project. reCaptcha was created to not only help secure websites from spammers/scammers/scummers, it was used to help with digitizing an ocean of old books being digitized using scanners and OCR software.

I imagine the tech has gotten better, but fifteen years ago, OCR software was only 80% or so effective at recognizing text scanned in from books. Humans have the ability to recognize and transcribe even blurred and distorted text with an accuracy of more than 99%, but hiring people sort out the 20% or so of the text scanned in was cost-prohibitive. We are talking projects that are attempting to digitize tens of millions of books.

So, what reCaptcha did was take chunks of unrecognized text that had been flagged and databased by the digitizing project and feed it to individuals trying to register online for various things. This required a minimal expenditure in time and effort for the people registering, but it has result in many hundreds of millions of previously unrecognizable chunks of text being correctly transcribed and it helped build ever more accurate OCR Algorithms. 
 

At one time, more than 40,000 websites used this verification system. I don’t know if the system is still in use by anyone, but it has done a tremendous amount of good in helping to digitize the collective knowledge and culture of Humanity, making it available to all.

[LostinTexas]

Alert the Village Elders at once!

[railfancwb]

[willie-pete]

Well, thanks for not making us do that to get in.

[Eric]

2 minutes ago, willie-pete said:

Well, thanks for not making us do that to get in.

I’m not doing that specifically right now, but other technologies are at work to keep the bots out. They are a freaking scourge, one that I have been battling for two decades. 

[Historian]

1 hour ago, Eric said:

 I don’t know if the system is still in use by anyone, but it has done a tremendous amount of good in helping to digitize the collective knowledge and culture of Humanity, making it available to all.

reCAPTCHA v1 was declared end-of-life and shut down on March 31, 2018.

I remember hearing an interview with the guy who invented reCAPCHA and even back then, at least five or ten years ago, it was an astonishing amount of work it did.

[willie-pete]

2 hours ago, Eric said:

I’m not doing that specifically right now, but other technologies are at work to keep the bots out. They are a freaking scourge, one that I have been battling for two decades. 

Yeah, they've got their hands full at the other place with that problem

[NPTim]

Wait! The robot makes you prove you’re not a robot? That’s profiling, and as we have rammed down out throats ad nauseum, profiling is WRONG!

[tadbart]

Scooters identifying as motorcycles? The gelding of humanity.

"Select all the boxes with taxis." Does that include that teensy little corner of the trunk of a taxi in that one box, or not? I don't think I've ever guessed wrong.

[railfancwb]

9 hours ago, Historian said:

reCAPTCHA v1 was declared end-of-life and shut down on March 31, 2018.

I remember hearing an interview with the guy who invented reCAPCHA and even back then, at least five or ten years ago, it was an astonishing amount of work it did.

I still get hit with those from time to time. Mostly when trying to access a linked Facebook item. 

[Huaco Kid]

So you're really a Rocker, and not a Mod?

[DWARREN123]

Not a robot, wife says I am not smart enough!

[Mrs.Cicero]

16 hours ago, Eric said:

One of the technologies used for this purpose was an interesting project. reCaptcha was created to not only help secure websites from spammers/scammers/scummers, it was used to help with digitizing an ocean of old books being digitized using scanners and OCR software.

I imagine the tech has gotten better, but fifteen years ago, OCR software was only 80% or so effective at recognizing text scanned in from books. Humans have the ability to recognize and transcribe even blurred and distorted text with an accuracy of more than 99%, but hiring people sort out the 20% or so of the text scanned in was cost-prohibitive. We are talking projects that are attempting to digitize tens of millions of books.

So, what reCaptcha did was take chunks of unrecognized text that had been flagged and databased by the digitizing project and feed it to individuals trying to register online for various things. This required a minimal expenditure in time and effort for the people registering, but it has result in many hundreds of millions of previously unrecognizable chunks of text being correctly transcribed and it helped build ever more accurate OCR Algorithms. 
 

At one time, more than 40,000 websites used this verification system. I don’t know if the system is still in use by anyone, but it has done a tremendous amount of good in helping to digitize the collective knowledge and culture of Humanity, making it available to all.

Oh, that's so cool!  I used to spend a few minutes each day working on a  project that was taking old letters/diaries/journals/handwritten stuff and converting it to something we could all read - since apparently cursive is not a thing anymore, lol - that's a neat way to do it faster.

[janice6]

Cursive will be in the next war, what native American language was in WWII.  "Code Talkers".

[Historian]

47 minutes ago, Mrs.Cicero said:

Oh, that's so cool!  I used to spend a few minutes each day working on a  project that was taking old letters/diaries/journals/handwritten stuff and converting it to something we could all read - since apparently cursive is not a thing anymore, lol - that's a neat way to do it faster.

Little Historian is learning cursive in school.

[Swampfox762]

1 minute ago, Historian said:

Little Historian is learning cursive in school.

Do they still teach that in public school?  I'm not bein funny, really don't know.

[Historian]

1 minute ago, Swampfox762 said:

Do they still teach that in public school?  I'm not bein funny, really don't know.

I wouldn't know.   I eat a lot of sandwiches to send that person to a fine Catholic private school.

That kid is about three years more advanced than i was at the same age.  But where i went to school the kids either went to jail, got married, or the military.

[RenoF250]

19 hours ago, Eric said:

One of the technologies used for this purpose was an interesting project. reCaptcha was created to not only help secure websites from spammers/scammers/scummers, it was used to help with digitizing an ocean of old books being digitized using scanners and OCR software.

I imagine the tech has gotten better, but fifteen years ago, OCR software was only 80% or so effective at recognizing text scanned in from books. Humans have the ability to recognize and transcribe even blurred and distorted text with an accuracy of more than 99%, but hiring people sort out the 20% or so of the text scanned in was cost-prohibitive. We are talking projects that are attempting to digitize tens of millions of books.

So, what reCaptcha did was take chunks of unrecognized text that had been flagged and databased by the digitizing project and feed it to individuals trying to register online for various things. This required a minimal expenditure in time and effort for the people registering, but it has result in many hundreds of millions of previously unrecognizable chunks of text being correctly transcribed and it helped build ever more accurate OCR Algorithms. 
 

At one time, more than 40,000 websites used this verification system. I don’t know if the system is still in use by anyone, but it has done a tremendous amount of good in helping to digitize the collective knowledge and culture of Humanity, making it available to all.

I have heard that before but it does not really make sense if it was ever designed to be a test.  If they don't know what the text is already how would it know you are not giving them a false answer.