Chapter 11: Counting and Compressing

Some stunts are about endurance of the machine. Others are about endurance of the person. In 2011, Jonathan tried both.

The math of counting out loud

Here is a question to ask a friend sometime: how long would it take you to count out loud from one to one hundred thousand? Most people guess wrong. They picture themselves rattling off numbers for an afternoon, maybe a long evening. The real answer is humbling, and it is the whole reason the next stunt needed two parts.

Try it small first. Count from one to one hundred out loud at a natural pace. You will notice something immediately. The small numbers are quick. One, two, three, four, they fly by. But the numbers get longer as they get bigger, and longer numbers take longer to say. "Seven" is fast. "Seventy-seven" takes noticeably longer. "Ninety-nine thousand, nine hundred and ninety-nine" is practically a sentence. So counting to 100,000 is not like counting to 100 repeated a thousand times. It is slower than that, because the words keep getting longer.

Here is the math, kept simple. Counting from 1 to 100,000 means saying out loud a total of 488,895 digits. Every one of those digits takes roughly half a second to speak, about 0.54 seconds apiece. Multiply it out and you get around 264,000 seconds of talking. That is roughly 73 hours. Three full days and nights of speaking numbers, without stopping, without sleeping, without a break.

Nobody can do that in one sitting. That is the first lesson of the stunt, and it is why "Jon Counts to 100,000" came in two parts. Part 1 alone ran for more than 74 hours, and even that was not enough. The numbers were not done. So he made a part 2 and kept going until the counting was complete.

Think about what that means as a human experience. Seventy-four hours is more than three days. You cannot stay awake for it. You cannot keep your voice steady for it. You have to sleep, eat, rest your throat, live your life, and come back to the camera again and again, picking up the numbers where you left off. The video is not a performance. It is a record of a person refusing to quit a ridiculous task.

What makes the counting videos work as a story is that anyone can feel them in their body. You do not need to understand video compression or server limits. You just need to have tried to count something long, and gotten tired, and kept going. The stunt turns a private childhood game, the kind of dare kids issue on long car rides, into a public marathon. It is the same dare, scaled up until the scale itself becomes the joke.

A minute inside the count

Try to picture one minute of it. Not the whole 74 hours. Just sixty seconds, somewhere in the middle, when the numbers are long and the novelty has worn off.

He is sitting in front of the camera. The number on his lips has five digits now, which means every number takes a full breath or more to say. Forty-one thousand, two hundred and eighteen. Forty-one thousand, two hundred and nineteen. Forty-one thousand, two hundred and twenty. Each one has to be right. If he loses his place, the whole project wobbles. So there is a rhythm to it, a pace he has to hold the way a runner holds a pace, except the marathon is in his mouth and his memory at once.

Now stretch that minute across a day. Across three days. The voice gets tired before the mind does, or the mind gets tired before the voice does, and either way the numbers keep coming. Somewhere in there he has to sleep. Somewhere in there he eats, drinks water, rests his throat, lives a normal hour or two, and then sits back down and finds the number again. The video does not show you a man counting. It shows you a man returning. Again and again and again, to the next number.

That is the human core of the stunt, and it is why the math matters. 488,895 digits at about half a second each is not a statistic. It is a description of what three days of a person's life sounded like. Every digit is a small decision to continue. Say it again: 488,895 decisions.

The slideshow: 571 hours in 549 megabytes

Earlier that same year, in July 2011, Jonathan tried a different kind of impossible. He took a trip's worth of photographs from Chile, photos of graffiti, dogs, stairs, airports, the ordinary scenery of travel, and turned them into a video 571 hours, 1 minute, and 41 seconds long. A photo slideshow running at one frame per second. If you had watched it straight through, you would have been watching for nearly twenty-four days.

This one was not about duration for its own sake. It was about a fight with the size of the file. A video is a file, and files have to be uploaded, and in 2011 YouTube had a rule: no file bigger than 20 gigabytes. Twenty gigabytes sounds like a lot until you try to squeeze 571 hours of video into it. A normal video runs about half an hour per gigabyte at modest quality. At that rate, 571 hours would have needed something like a thousand gigabytes. The cap was twenty. The math did not work. Not even close.

So Jonathan did what he had done with the two-tone video: he packed light. He shrunk every photo down until each image was a tiny preview roughly eight pixels by eight pixels, closer to a mosaic tile than a photograph. One frame per second, tiny frames, and the simplest possible video. The whole 571-hour slideshow squeezed down to 549 megabytes. Half a gigabyte. That is less than three percent of the 20-gigabyte limit.

To picture what that squeeze meant, imagine the upload cap as a suitcase. YouTube says the suitcase can hold 20 gigabytes and no more. Jonathan had 571 hours of travel photos to fit inside it. If he packed them at normal size, he would have needed fifty suitcases. Instead he folded everything down until it was nearly flat, pressed out all the air, and fit the entire trip into a corner of one suitcase. The 549-megabyte file is the feather in the suitcase. It is almost all air.

And here is the lovely contradiction at the heart of it. The slideshow is a record of a trip to Chile, a place full of color and detail, and the version that went up on YouTube was reduced to images eight pixels wide. The graffiti, the dogs, the stairs, the airports: all of them present, and all of them blurred down to near-abstraction by the demands of the file size. The trip survived, but only as a ghost of itself. That is the price of the stunt, and also its poetry. He kept everything, and kept almost nothing.

What 8 by 8 pixels actually means

It is hard to picture how small eight pixels by eight pixels really is, so let us build it up from zero. A pixel is the smallest dot your screen can show. Everything on this page, every photo you have ever seen on a screen, is made of pixels lined up in a grid. A modern phone screen holds millions of them. A single letter of the text you are reading right now is probably drawn with hundreds.

Eight by eight means sixty-four pixels total. Sixty-four dots to draw an entire photograph. A dog becomes a brownish blur in a greenish field. Graffiti becomes a smear of bright color. An airport becomes a gray rectangle with a lighter streak. You could cover the whole image with your fingertip and still have room to spare.

Why shrink them that far? Because every pixel costs file space, and he had over two million frames to store. At one frame per second, 571 hours is roughly 2,055,000 individual images. Two million photos, each one sixty-four dots. Even so, the video that came out of it was 549 megabytes, which tells you how tight the math was. If he had kept the photos even a little bigger, sixteen by sixteen, say, the file would have ballooned past what the plan allowed. Eight by eight was not a style choice. It was the size the puzzle demanded.

There is an honesty to this that is easy to miss. He could have cheated the stunt. He could have used fewer photos, or looped them, or faked the length. Instead he did the real thing: every photo from the trip, one per second, for 571 hours, each one honestly shrunk to fit. The file is small, but nothing was skipped. That is a very Jonathan kind of integrity. The stunt is absurd, but the execution is dead serious.

The same hands, different puzzles

Put the three stunts side by side and a pattern emerges. The two-tone video asks: how long will the machine allow? The counting videos ask: how long can a person keep going? The slideshow asks: how small can 571 hours be made? Three different puzzles, one set of hands. And all three were solved in the same year, 2011, the year he graduated from college with a degree in television production and turned, immediately and fully, to testing what the internet's new canvas could hold.

It is also worth saying what these stunts are not. They are not pranks. Nobody is being tricked. They are not stunts in the daredevil sense, where the danger is the point. They are closer to experiments, run in public, with the results posted for anyone to check. The lab notebook is the video itself. That public-experiment spirit is the thread that connects the 596-hour video to everything he made before and after. The kid who filmed a talk show in his bedroom was always running experiments on an audience. The stunts just swapped the audience for the platform.

The invisible work: rendering and uploading in 2011

It is easy to read about these videos and picture only the clever idea. But every one of them had a second, invisible phase that took real patience: making the actual file, and getting it onto the internet.

Rendering means the computer builds the final video, frame by frame, from the raw material. A normal home video might take a few minutes to render. A 571-hour video means the computer has to assemble over two million frames, one after another, without crashing, without a power outage, without the program giving up. On a home computer in 2011, that was not a quick errand. It was an operation you started and then left alone, hoping the machine would still be standing when you came back.

Then came the upload. Even squeezed to 549 megabytes, the slideshow was a large file for 2011 home internet, where uploads crawled compared to downloads. And the 596-hour video, light as it was, still had to travel across that same slow connection. Uploading meant the computer stayed on and connected for hours, sometimes days, with the progress bar creeping forward. If the connection dropped, you started over.

This is the part of the stunt era that deserves its own quiet respect. The ideas were bold, but the follow-through was pure grind. Anyone can imagine a 596-hour video. Finishing the render, nursing the upload, watching the progress bar for a day and a half: that is where the stunt actually lived. The dare was issued to YouTube, but the labor was all his.

An honest note about what survives

This book is built on records, and it is honest about the gaps. As of this writing, the 596-and-a-half-hour video and the Chile slideshow are not on YouTube anymore. They came down with the rest of the early channel history, and what remains are the records: the titles, the lengths, the upload dates, his own accounts of why he made them, and the documentation gathered since.

That matters, and it should be said plainly. A coffee-table book about videos is, in part, an archive of things you can no longer click on. Where the videos survive, this book points you to them. Where they do not, the book says so, and tells you what is known from the record. The counting reference numbers are solid: 488,895 spoken digits at about 0.54 seconds each, part 1 running past 74 hours, a part 2 finishing the job. The slideshow numbers are solid: 571 hours, 1 minute, 41 seconds, one frame per second, squeezed to 549 megabytes under a 20-gigabyte cap. The moon-shot quote is his own, from his own mouth, in 2021. The challenge is his, in his own words.

The stunts were always about the attempt more than the artifact. You do not need to watch 596 hours of two tones to get the joke, and you were never meant to. The attempt was the art. The record was the point. And the record stands, even where the files no longer do. Somewhere in the history of YouTube, there is a line that reads: this is how long a video could be, and this is the young man who found out.

Why 2011 mattered

Step back and look at the year as a whole. In 2011, Jonathan graduated from college with a degree in television production. He had spent his whole young life learning to make TV the old way: cameras, tapes, editing software, burned discs. And the moment he finished school, he turned to the newest canvas available and started testing it the way he had tested everything since he was thirteen: by filling it completely and seeing what happened.

Three stunts, three questions, one year. How long will the machine allow? How long can a person keep going? How small can 571 hours be made? Each one is a complete thought, and together they are a kind of thesis statement. This is what he does. He takes the whole canvas.

The stunt era did not last. It was never meant to. It was a burst, a young maker firing off everything he had learned at the exact moment the platform was new enough to be surprised. What came next was slower and stranger: ten thousand silent videos of drinking water, one after another, for over a decade. But the stunts were the overture. They announced the themes. Duration. Discipline. The whole canvas, filled. If you understand 2011, the decade that followed starts to make sense.