April 7, 2026

As I get older, I have found myself looking for beauty in the mundane rhythms of everyday life. Maybe it is because I have come to peace with the fact that life will be a lot more repetitive from now on, or maybe it is driven by a quieter belief that there must be something more to it than what is immediately visible. Regardless, I have grown an affinity for examining ordinary actions and trying to understand whether they carry some deeper meaning, or whether they can be repurposed into something larger. In a sense, I am trying to find meaning in what appears, at first glance, to be meaningless.

It is within this frame that the idea of “human computation” becomes particularly compelling. The concept, developed by Luis von Ahn, is deceptively simple. Instead of viewing human effort as something that must be explicitly organized and compensated, it proposes that small fragments of cognition, perception, and recognition can be captured within everyday interactions and aggregated into something of broader value. The individual act remains trivial. The system it feeds into is not.

In my opinion, Luis von Ahn, who is most widely known for founding Duolingo, is one of the greatest, yet most overlooked, technologists of our time. While Duolingo is an exceptional product in its own right, I am particularly fascinated by his earlier work on reCAPTCHA.

At a surface level, reCAPTCHA functions as a security mechanism, distinguishing humans from automated scripts. It evolved from earlier CAPTCHA systems, which presented users with distorted text that was easy for humans to read but difficult for machines to interpret, creating a simple but effective barrier against bots. But its structure reveals something more intentional. Early versions of reCAPTCHA presented users with words that machines had failed to recognize when scanning books. Each prompt typically included one known word, used to verify the user’s accuracy, and one unknown word that required transcription. When a user entered both, they were not only confirming their identity but also contributing to the digitization of text that machines could not yet interpret.

When Google acquired reCAPTCHA in 2009, the same logic was extended. Instead of text, users were asked to identify objects in images, traffic lights, crosswalks, storefronts. These were not arbitrary prompts. They were aligned with the needs of improving computer vision systems. Once again, an action that takes only a few seconds became a unit of data, contributing to systems far removed from the immediate interaction.

This was not framed as participation in a larger project, nor did it require any awareness. The interaction was already necessary, the system simply gave it an additional function. At scale, these small, routine inputs became a continuous source of labeled data, quietly contributing to one of the largest infrastructure efforts of our generation: training the AI models we now use in our everyday lives.

And this is where the idea of beauty in the mundane becomes real.

On the surface, typing in a CAPTCHA feels meaningless. It is a brief interruption, something to get through so you can move on. There is no sense of contribution, no indication that anything of value is being created. But beneath that surface, millions of those same actions were being aggregated into one of the largest datasets ever assembled, quietly helping power the development of modern AI.

The action itself never felt meaningful. But it was.

If we only evaluate our lives at the surface level, we miss this entirely. The mundane will always appear empty if we do not look beyond it. But when connected to something larger, even the most repetitive, overlooked actions can carry weight.

In our everyday lives, we are constantly contributing to systems far larger than ourselves, often without knowing it. Discovering and learning about these systems keeps me sane, as it helps me fully understand the beauty in the mundane.