Gen AI has a cool problem
And it won't be fixed by Elon Musk memes
Generative AI is not in the least bit cool. I say that as someone with three kids and a mortgage. So any vestiges of cool I had, and there wasn’t much to start with, are long gone. But that makes this even more damning! Even I know generative AI isn’t cool.
But everyone is talking about generative AI, that’s cool, isn’t it? That could be cool, if the conversations were different. But online, lots of these conversations come through the prism of money. People (not you, I’m talking about other people) are hoping for the next digital gold rush. They look back at inflection points around anything from crypto to vlogging to influencing and think “what if I had gotten in at the ground floor?” That informs a large chunk of the discourse around generative AI.
But it’s not just a profit motive, there’s also the push of fear. That if you don’t engage with these innovations you’ll be left behind. If the robots are taking our jobs, maybe you can at least volunteer to be their assistant. Whether it’s fear, or ambition, the motivation is linked to productivity, efficiency, all sorts of boring things that are deeply unsexy at best, outright dystopian at worst.
That’s not an environment that particularly attracts creativity. And many creatives are already likely to feel an immediate ick when presented with generative AI possibilities. It feels at odds with the personal element of the process that they hold dear. So there’s a drastically smaller pool of creatives exploring what might be done with these tools. The scene, if a scene is even an applicable word now, is dominated by fan-art and misogyny. There’s not been a breakthrough piece of media that feels counter-cultural.
How big a problem is it? Not massive, yet. Can it be solved? Possibly.
It’s OK, keep subscribing
Today’s issue will focus on the cool problem. But on Thursday we’re going to curate some of the content that actually is pushing boundaries. Is there anything you have seen over the past year that stopped you in your tracks? Get in touch on contact@explainable.online
Evidence for the prosecution A: The Elon AI problem
Tiktok failed to load.Enable 3rd party cookies or use another browser
Elon Musk is a hard person to avoid online. For the past 18 months he’s been at Trump 2016 levels of ubiquity, reaching that peak (or trough) where haters and supporters are equally happy to post about him. And that coincided with the gen AI boom and the rise of the Musk-era blueticks on Twitter (for the moment, we’re sticking with Twitter in the Explainable style guide) which means a lot of fan art of Musk as, say, a gladiator or a grizzled mechanic or an inspo bro.
The fan art comes from a mix of genuine admirers, people who like using phrases like “triggering the libs” and very online folk who are multiple levels of irony down on this and may not even themselves know their true feelings on the guy. It’s gen AI as the latest shitposting tool. It’s everywhere and it’s weird and sometimes funny and sometimes baffling. But it’s definitely not cool.
Evidence for the prosecution 2: TikTok and AI
Tiktok failed to load.Enable 3rd party cookies or use another browser
It’s debatable if TikTok is cool anymore. The brands have long since arrived and the boomers followed them shortly afterwards. But it’s still the most likely place for a truly innovative piece of video AI to pop, with YouTube more likely to host the how-tos of AI tools; rather than the finished product.
But pickings are slim. Plenty of AI tool promos, some basic how-tos and lots of AI filter dancing trends, with some Will Smith eating spaghetti thrown in just to keep things vaguely interesting. But there are few examples of AI being used to create something that hasn’t been seen before. The popularity of TikTok was rooted, as much as anything, in how it made basic video editing tools accessible to users who didn’t want to spend a day in Premiere Pro. So the AI filters are a continuation of that, offering broad easy-to-understand tools in-app.
It’s democratized video creation, but it’s worked against the people who actually like spending a day messing around with Premiere Pro. In peak YouTube vlogger days even the cool kids had some AV club geekery in them. The people going deep on AI tools aren’t breaking through, yet, on TikTok.
Case for the defense
All online creativity has a basic timeline of:
Engineers badly demonstrate new technology for an audience they don’t really understand
Early adopters over-represent their niche interests
Geeky kids who grasp the tools, while also kind of knowing what ‘it’ is, offer innovation
The geeks go mainstream and the cool stuff is recycled and reiterated
Enter: the brands
And that timeline has gotten more compressed every time as influencers and brands learn from the past and try to anticipate new developments. The paradox with generative AI is that everyone expects immediate seismic changes. Most discourse around AI is about how quickly the sands are going to shift under our feet.
But in reality, at least when it comes to creativity and these tools, we’re back to an early, slow-moving stage of the above cycle. Engineers are badly demonstrating tools, and early adopters (read: teenage boys) are focusing on niche interests (read: fantasy and questionable depictions of women).
Rather than make everything super easy, gen AI has just provided a new group of tools to learn. The brands and influencers are still hovering, desperate to jump on board, but they haven’t seen that important middle stage yet. Where cool, innovative art gives the wider public an “oooooh, I get it” moment.
AI in the wild
To get away from doom and gloom at the last. This post did numbers last week. The image is from an entertainment news account with an AI images-only hook. They post enough bait to suggest this was deliberate, particularly given that it remains up. The “look at what the stupid AI did” content cycle has many spins left.