The Tour de France is happening right now and every now and again people who know I don’t like most sports and also that I like cycling assume that that is the sport I follow. It isn’t and for me it’s baked into its structure. People are hyper-optimized both legally and illegally and supported within an inch of their lives. The goal is one thing: come in first. And for that I can’t get invested. I do follow some of the ultracycling races you can “watch” online at places like dotwatcher.cc. These ones are often unsupported, sometimes letting riders figure out their routes themselves like on the Transcontinental Race across Europe where only certain segments are mandatory to ride – the rest is up to you. I have a little bit of interest in who comes in first but often I’m more interested in the folks who are sharing their experience online as they’re racing. They often are having very human experiences: mechanical failures, frustration, joy at the beauty of a vista, a missed route or a delicious lunch. This, to me, is a relatable and interesting human experience. The first feels, to me, like it could just as easily be performed by autonomous robots. Perhaps in a few years it will.
But what I also notice about the two is that the motivations of the two seem very different. People want to win the Tour de France often at all costs, even to cheat. There are certainly benefits to winning an ultracycling race but the focus is often on others in the race as well – personalities as well as performance. Folks like Han Lu who fought through tons of challenges, finishing the Transcontinental last but making sure she did finish it. You can hear her story here.
In other words, I’d rather watch this clip of Lael Wilcox being encouraged by a bunch of schoolkids on her world record circumnavigation attempt:
Than this “exciting” TdF clip.
I think it’s because for me my connection to the content I consume is about story and emotional connection.
And that distinction is what comes to mind when it comes to so-called art and writing created by generative AI. Yes, it’s passable – it can be a picture or tell a story and those things are getting increasingly better. And if that’s what you’re looking at art for: “Look at the pretty picture” then it will serve you well. If you’re looking for humanity, keep looking. It may seem human but at it’s heart it’s full-featured autocomplete. Or as one person put it: “It’s a plagarism machine made of crimes”. A good cyclist might be able to be win the TdF with a few extra supplements or illegal bike modifications but in the end, that winner is not entirely human. That story submitted to a literary magazine? It wasn’t a human story.
As long as there is motivation for some to take credit for “enhanced” or fully plagiarized work whether for fame or money, there will be folks trying to pass off others’ effort as their own.
So what to do?
In the academic world I have heard of the arms race between students submitting AI-generated assignments, AI-detection software, and then “humanizing software” to beat the detection software. Meanwhile I’m over here questioning the whole process. If we are trying to teach how to do tasks that are now going to be AI-assisted or fully AI-performed then that’s likely not worth the time at all – or re-frame the assignment to be AI-inclusive.
If we’re trying to ensure that students have knowledge and understanding committed to memory then our methods are inadequate. Many are already moving back to in-class assignments and activities. This NY Times article (paywall free link) does a great job of describing this. It feels like in some ways it’s resulted in an improvement.
In short form video on places like Instagram and TikTok I’m seeing artists not only showing what they created, but many videos are now showing not just the output but the whole creation process. To me this is more interesting than the content alone.
In the end, my reaction to AI is more about questioning what the demand is for. What are we hoping to get from the content we’re consuming? Are we looking for a beat to dance to? Sure, AI can do that. But if you’re looking to hear someone sharing their thoughts and stories that leave you wanting to chat with the artist and hear about the origin of it? Maybe an AI could create this:
or this piece that Sage did, but would it mean as much to you knowing that a computer put it together as a sequence of sounds likely to make you happy instead of wanting to share something with you?
What about you? Would a fully AI-generated blog, or AI-enhanced blog (maybe it takes my week’s camera roll and a few voice notes and turns it into an entry) be the same for you?
Featured image source: Local Artist, Christopher Reid, Painting Chapman’s Peak Drive by Kimberly Reid Creative Commons Attribution-Share Alike 4.0
I love your analogy about the journey, where some people want to win and some people just want to see if they can. I see a lot of similarities with the writing field right now.
I’m currently working at the intersection of marketing, technical writing, and personal writing. Personal writing keeps me sane, but it is my professional writing that still slightly discombobulates me.
I see a coming trend where marketers will write for LLMs, structuring content in a specific way so that LLMs will cite them. Already, writers in marketing are feeling the pressure to do this. You really want to write like a human being, but LLMs want you to structure content in a certain way. I attended a class and was shown how to do this. At the end of the class, I said, “As a marketer, I’m really excited for what I’ve just learned. But as a writer—one who has been doing personal writing, like essays and fiction, for over 30 years—I’m a little bit horrified.”
In the past, we worried about SEO content and got upset by how content had to be written for SEO. In the future, content will be written for LLMs, and that is pretty upsetting.
The only silver lining I can see is that there are no AI monopolies right now. When we wrote for SEO, because Google is a monopoly, all content started sounding the same because you had to write for one provider. Right now, between open-source models and the commercial ones in the United States, there is a lot of diversity. It is pretty hard to write for a specific model or to gamify it, so to speak. Because of that, there will at least still be some room for a writer’s voice.
But it is still upsetting that people in the marketing field want to be like those riders who want to win at all costs, meaning they will take as much support and supplements as they can get to do so.
I agree with you Todd. Anything AI written is not from the heart . When we write there are mistakes but they make the writings more acceptable. I agree about the cycling and I feel the same about cricket. Yesterday we watched the Wimbledon ladies final. It was definitely not a mechanical game. We too got involved and I like to watch the emotions on the person’s face and body language after winning at Wimbledon. Thank you for this post.