I’m Not A Robot, I’m Working Without Knowing It

I don’t know, maybe I’m thinking about this too much, but some things really make me stop. For example, when I clicked “I’m not a robot” or when I read those crooked, twisted letters and typed them into a box, what was I actually doing?

When I learned the answer, I felt something inside. Not exactly anger, not exactly admiration. Something strange in between.

How CAPTCHA Was Born

Let’s go back to the beginning. In the early 2000s, Luis von Ahn, a professor at Carnegie Mellon, yes, the same founder of Duolingo, learns during a talk that Yahoo is in trouble: bots, meaning written programs, are opening thousands of fake email accounts. The computer cannot tell humans from bots.

Von Ahn thinks about it and finds the solution: at the time, computers cannot read distorted, crooked text, but humans can read it easily. So let’s use that as a test. This is how CAPTCHA is born. The test works, everyone starts using it, and it is entered 200 million times a day.

200 million times. Every day. Just to say “I am human.”

While I Was Proving I Was Human

And then Von Ahn has a second idea. This is the part that really affected me. The New York Times and Google were digitizing old books and archive documents in those years. But scanning software could not correctly read blurry, crooked characters. Right there, a very elegant solution enters the picture: they show these unreadable images to people as CAPTCHA. The human reads and types them, both proving they are not a robot and taking on the digitization work at the same time.

So what happens, in short?

While you are trying to prove that you are not a robot, you are also manually typing The New York Times archive or the Google Books database for free.

Maps, Signs And Invisible Labor

A similar logic also enters the picture on the Google Maps and Street View side. How did you feel when it said, “Select the squares with traffic signs”? I usually thought, “God, how annoying.” But at that moment, we were becoming part of a process that helped machines better understand objects, texts and signs in street images.

And Von Ahn continued this approach for a period in Duolingo too: users translated real news articles while practicing. Agreements were made with BuzzFeed and similar sites, and their content was presented as Duolingo exercises. You are happy because “I am learning,” while at the same time becoming part of a professional translation service.

Clever Or Disturbing?

When I learned this, my first reaction was “wow, good thinking”, I admit it. You rarely see pragmatism in such an elegant form: a security problem is solved, human labor becomes useful, and the party benefiting from the service does not even realize it.

But on the other hand, one question stayed stuck in my head: millions of people worked for years without realizing it. Digitization, labeling, translation... Who paid the price? Who made the profit?

This question may look more elegant in the digital age, but labor being used without being noticed, minimized, or explained differently is not new. For a much darker and more physical example of this, The Radium Girls: American Workers Who Died Making Glowing Watch Dials >> looks at the same disturbing question from another place.

Maybe from the system’s point of view, it is a perfect loop. But when you look at it from the human side, it leaves a strange feeling. As if someone said to you, “please enter through this door,” and you entered, only to carry a load without realizing it.

I Don’t Look At The Checkbox The Same Way Anymore

Did this make me feel bad? Not exactly. But now, whenever I see a CAPTCHA, I pause for a second before clicking the checkbox.

“Whose work am I doing right now?”