All hail the rise of the machines
The rapid pace of technological change can be both exhilarating and terrifying, leaving many of us feeling lost and unsure of how to navigate the rapidly shifting landscape. As new technologies are developed and old ones become obsolete, we are forced to adapt and learn new skills in order to keep up. But this constant change can be overwhelming, leaving us feeling disconnected from the world around us and unsure of our place in it. The machines that were once tools to help us now seem to be racing ahead, leaving us behind. And as we struggle to catch up, we can’t help but feel a creeping sense of dread about what the future holds and whether we will be able to keep pace.
The previous paragraph was entirely written by a machine. I asked it to write something haunting about the pace of technological change, and this is what it produced. I have the eye of an editor and the soul of a writer, yet I can’t find much to critique in that passage. It is grammatically perfect. It’s a deep, complete thought—one that could have been pulled from any technology article in WIRED. The sentences are conversational and complex, including phrases like “both exhilarating and terrifying” and “feel a creeping sense of dread.” I could not identify this writing as anything other than human.
“You will see fake images and you will not know the difference.”
But it isn’t human. It’s not even exactly intelligent—the program that generated that text is a “language model” called ChatGPT. Language models scrape huge amounts of text data from all across the internet and scan it for patterns. They are then carefully trained by humans to produce responses to text prompts that are relevant, factual, and well-written. In other words: ChatGPT read a ton of stuff, looked for language patterns and facts, and now it can spit bullshit back to you in perfect English. It’s not thinking about your question; it’s mashing together bits of all text it has ever analyzed and then forming a plausible answer from that soup.
This magazine is about snowboarding. I love snowboarding because it takes me up into the mountains where the cold air sharpens my focus and clears my thoughts. I don’t ride with music anymore because I want to hear the snow and the wind. Shredding is an action that takes place completely in the moment, at speeds humans rarely experience, and it requires balance, dexterity, and high-consequence split-second decision-making. We do all of this automatically, on every run, without thinking. Snowboarding is a marvel that only human beings could achieve.
The photos accompanying this story were also generated using an AI tool. The process is similar: enter a text prompt, and the program produces an image based on millions of other photos, illustrations, and artworks that it has been trained on. The results are occasionally nightmare-fuel: grotesque, cursed images of twisted humans with backwards limbs and horrifying grimaces. But more often than not, the program will produce a facsimile that’s hard to distinguish from the real thing. It can mimic any artist and any art style, and can produce everything from abstract works to photos. Not just images that are photo-realistic: shots that are indistinguishable from actual digital photography, apart from small digital artifacts and errors. On a phone at Instagram-size it can be impossible to tell they’re not real.
What happens when these tools get so good that we can no longer tell the difference between a real photo or video and one created by an AI program? We have built the entire economy of snowboarding–and the entire economy of the world–on the commodity of human attention. I’m already seeing the limits of that saturation; with so much media screaming to be heard, I find myself no longer interested. That saturation is going to get exponentially worse.
When I scroll through Instagram or TikTok I’m presented with content featuring other humans doing things. These people are creating that content in the hope that I will interact with it, the algorithms will then favor their content, and they will be rewarded with more attention and, maybe, money. I trust that what I’m seeing really happened, but I fundamentally have no way of knowing that for sure. The day is not far off when people are going to start going viral for clips and photos created by AI. The attention economy will guarantee that they will be rewarded for taking this shortcut. You will see their images of athletes hitting backcountry kickers and making sunset pow turns, and you will not know the difference.
This will break the attention economy. When any piece of media could be fake, without any way to tell, all media will be worthless. You might as well be watching a Pixar movie rather than a snowboard film. Hell, at that point maybe Pixar movies will also just be regurgitated automatically—an endless feedback loop of AI-generated plots and characters. Eventually the machines will just be copying themselves.
Now there is a haunting paragraph about the pace of technological change. Using these tools can be exhilarating: There is a wild joy in how mind-blowing it all is. And I can’t predict the future with any kind of accuracy, only speculate on the eventual end result. What I do know for sure is that computers can’t snowboard. They can’t stand pressed against a frosty Tram window watching the cliffs on either side of Corbet’s materialize out of the morning clouds. The only experiences that can’t be manufactured are the ones you have yourself.
When you go riding with your friends, that’s as real as it gets.
– OL