Close reading the trees

How competitive Google Street Viewing makes the world feel seen

This week’s guest post comes from my dear friend, Adriel. If you think that Adriel should publish more (as I do), vote with your email address by subscribing here.

---

I'm doomscrolling late at night when I come across my first Rainbolt video. He’s guessing where in the world pictures of completely nondescript dirt roads are located, and he's guessing correctly. It's a show of knowledge that strains credulity, and it's too late at night for me to start investigating. Can someone really tell the difference between a dirt road in South India and a dirt road in Brazil without other identifying features? I throw him a follow, determined to come back to this at another point and verify that it isn't a prank or gimmick. To my happy surprise, it's much more. 

georainbolt

A post shared by @georainbolt

Rainbolt started out competing in an internet game called GeoGuessr. On GeoGuessr, you are shown a random Google Street View image, and you have to guess where in the world it was taken. The closer to the real location that you drop the pin, the more points you get. Rainbolt is in the top echelon of players, and to say that he can easily locate a screenshot is an understatement. He tries his best to hinder himself, sometimes blurring the pictures, or making them flash for only a fraction of a second on the screen, to show just how thoroughly he knows the map. He does this while remaining cool, calm, and understated. "Nice," he characteristically says to himself, after guessing the location of a field of grass within 100 kilometers of its true location in Botswana. You get the sense that he holds himself to a high standard. 

There's a way in which the internet stifles creativity, boxing people into their original creative moment and fifteen seconds of fame. When people become internet famous, they tend to trap themselves into the bit that made them viral. But Rainbolt’s skills evolved in new and interesting directions. People started sending him old pictures, sometimes with loved ones who had since passed, and Rainbolt helped them identify the location of their photos. Sometimes he took music videos, or memes, or whatever else was meant to be background and bland and universal, and exposed precisely where the film was shot or picture was taken. He can take the most banal backdrop and make the world less flat. 

I don't think that's trivial, because lately I have been feeling that the more we travel and the faster we communicate, the more indistinguishable the world becomes. This feeling crystalized for me in my late 20s when I got on dating apps and found them shitty. They were shitty in a lot of ways but one of the most apparent to me was the steep education in signaling, the way you quickly learn to decode the universal dating app language of cultural and tribal affiliation. At the top of the pyramid were the terrible same-y travel cues: pose with an elephant, pose on a beach, pose in front of a landmark, Instagram highlights categorized by airport code. Hinge prompts were somehow worse “What do you most look forward to this year?”… “travel!” “I want someone to”… “plan my next trip with!” Travel was supposed to function as a symbol (disposable income with a veneer of sophistication, perhaps), and as the symbol ate the underlying meaning, the meaning became bland: restaurants in foreign cities identical to the ones here in New York, generic skyscrapers barely recognizable as from anywhere, Coldplay in an open air market in Hanoi. Somehow Rainbolt feels like the balm against all of this. He can’t reverse easy and empty status signals, but he can at least remind us how textured the world is. Even in the least distinctive photo, there are signs of specificity if you know where to look. 

Now, I know... if Rainbolt can do... shouldn't we be able to train a computer to do this as well? I'm not knowledgeable enough in the technicalities of what that would take, but based on recent developments in AI, I'd be surprised if you couldn’t. If I can imagine it, I suppose someone must be building it [editor’s note: they are]. We are probably more unsafe than we know on the internet. 

But for now there are moments of stunning humanity, a short term triumph even, of human over computer. Because what we are witnessing here is a reversal of the mold, in which humans input a massive amount of data into a computer that outputs something new and searchable. Here the computer is the input that the human scans, learns from, trains on, eventually outputting something magnificent. Because Rainbolt is just a human, we don't have to worry about the sad and bizarre and freaky edge cases. The humanity shines, whether through humor (schooling some rando who thinks they are "off the map"), love (locating the last place someone got a photo with their dog), or ego (proving he can also ID places if the picture is upside down and pixelated). His posts are never uncanny valley.

Lately, Rainbolt has been traveling the world, and is letting us in on the secrets he has seen in the data. Did you know that square stone fences with circle holes are only found in Chiang Mai, red stop signs with white font are only found in New South Wales, and the Wooden poles are cut in an octagonal shape only in Sydney? It's exciting to see him get off the computer and traveling the world in real life. Rainbolt is a modern explorer, an internet savant, and he’s been training at home so that his trip can be even more interesting. Maybe the internet won't just swallow us whole. Maybe we'll use it to appreciate just how distinctive humanity really is. In the meantime you can follow him on Instagram, and when the doomscrolling gets too much, maybe he’ll bring you back to the ground. 

---

Thanks to Adriel for writing! Subscribe to his newsletter here:

---

🐌 Snail mail:

Last week’s piece about Substack’s dubious network effects sparked a chorus of head nods from other writers who were wondering if it was all in their head, or had their own observations/insights to add.

From

Sarah M. Chappell

:

My anecdotal experience matches yours. The over-reporting tracks for me personally. Only the smallest sliver of new readers have come through Substack. Are there network effects? Sure. Are they remotely evenly distributed amongst comparable publications or approaching the level that Substack promotes? Nope. I've worked with online community/creator/small biz etc. for many years and the host platform is very, very rarely the thing that makes a successful product. Like Patreon before, Substack has a massive incentive (10% worth) to tell that story, but I don't believe it's true.

From

Anna Grigoryan

:

the world of "network effects" is different for small newsletters compared to large newsletters. In case of many high profile newsletters like Lenny's Substack itself is putting great deal of work into promoting them. They [the newsletters] are featured on Substack's platform. The writers frequently give webinars to other writers, they are active on other social media. So those "network effects" for large newsletters are just that THEIR network effects rather than Substack’s.

From Sylvia:

I happen to have just commented to someone that I’m nervous about leaving because of the readership that Substack has brought me, exactly as you predicted. I've now set up on Beehiiv and intrigued to see what happens if I promote that link while still remaining active on Substack.

---

Welcome to our new readers, some of whom described their relationship with The Algorithm in one emoji thusly:

🦡 🎷 👀 🕌 🌇 👽 🦝 🪸

If you had Rainbolt’s super powers, how would you use them? I for one have always wished I was the kind of person who could locate myself based on the fauna.

---