A bunch of speculative money in the world of AI was lost this week. Boo hoo?
But it's also undeniable that AI, a term that has in many ways lost any and all meaning, will be a major part of the coming years. Its loss of meaning matters not. It might very well go the way of crypto, but for the moment, it's inescapable.
How do you handle your children interacting with AI videos? What will it mean for young people to enter an internet full of lies and, as the headline implies, slop?
Naturally, Rob and Patrick have some thoughts.
Patrick: Rob, it was a little strange to watch the headlines this week about billions in market capital disappearing and simply blink. It appears the next century will not necessarily be dominated by Silicon Valley guys who love to align with fascists while chatting on podcasts!
Who could’ve thought?
I blinked because it’s been strange to watch the “AI revolution”—an endless, accelerated stream of hype, money, discourse—and feel so utterly disconnected. Besides ire at its obvious potential and material impacts on labor and art, it often feels like a situation where I’m being told my life is going to change. No. If you buy the arguments, has changed. The way Apple pitches its ludicrous Apple Intelligence makes it sound like the world is already shifted, when in reality what they’re pitching are notification summaries that mostly suck. I’m trying to figure out the FOMO!
I have, on occasion, asked ChatGTP a basic history question out of curiosity. It seemed fine. Seemed fine. But I lump ChatGPT in the same bucket as the dreadful “AI” summaries that you find while doing a Google search these days. Consulting ChatGTP about a sports question, though, is a choice. Google’s AI summaries are not, and busted as they might be, I have found myself defaulting to scanning them for information because they take up the most important real estate that Google has: the first thing you look at when a search has gone through. I hate it.
Can I tell you an upsetting story?
My eight-year-old is in a theater class at the moment and their production is The Wizard of Oz. (My child, bless her heart, swung for the fences to be Dorothy in auditions, but it’s a class for older and younger kids, so she’s a munchkin. Thankfully, she’s very excited to be a munchkin.)
While picking her up this week, she excitedly asked me “Dad, did you know there’s going to be a sequel to The Wizard of Oz?” I blinked, figuring she meant the second Wicked movie that’s coming out this year. All we’ve done lately is listen to the soundtrack and watch inspired dances on TikTok. But before I could ask, her friend came up and gave more details about this sequel.
Apparently, this sequel is coming soon “according to a trailer on YouTube.” I’m not sure if you’ve made the mistake of accidentally clicking on the ludicrous amounts of fake trailer AI slop on YouTube, but once you make the mistake once, your algorithm is fucked. It just swarms you with AI content from then on. I wasn’t going to ask this 10-year-old if what they’d watched was AI or not, but there’s little doubt in my mind they watched some AI video slop and took all of it literally.
I quietly sighed internally, told the kids “oh, that sounds cool,” and walked out.
It’s going to get so much worse!
Rob: That’s the vision Silicon Valley has for us, isn’t it? Promise a massive hardware and resource-intensive infrastructure to power a wholly owned counterfeiting machine, because the sheer absurdity of those demands justifies raising impossible amounts of capital for investors who want big, easy yields on their already ridiculous piles of assets. Use the counterfeiting machine to create content for platforms that have an insatiable appetite for audience size and engagement numbers. Trick elders and children at scale, but if that strategy doesn’t produce sufficient growth, create entire fake “content creators” reaching audiences with increasing numbers of bots performing bleak pantomimes of online life. The end result is theoretically a thriving economy of black box platforms claiming growth that is impossible to verify, blotting out the ability of real creators and audiences to find one another or even be certain that both parties are actual human beings. Is there a real hardware innovation here? Hard to say, it seems like the OpenAI model was partly predicated on being able to buy more expensive hardware than anyone else.
Is there platform innovation? If fraud is innovation (and investors largely decided those two things were synonymous since 2000), then it might qualify insofar as it solves the problem that Facebook’s membership is aging and disengaging by replacing real, live human users with machines posting a slurry of AI content into their feeds. The plans Google and Facebook have for your kids involve them spending their lives glued to their phones watching AI-generated short-form video that makes them feel poor, unattractive, and insufficient, because that is the ideal consumer on those platforms, assuming all their money doesn’t end up in Roblox.
Now this week’s big news is that a Chinese-funded and developed competitor to Silicon Valley’s LLM boondoggles called DeepSeek appears to be able to significantly outperform their programs at a fraction of the cost and hardware resources. If all this pans out, it’s possible that a lot of Silicon Valley investors are going to lose massive amounts of money and will not own nearly as much of the infrastructure of this AI dystopia I described above. However, a lot of the companies who want to use AI against their users and workers will not care where the models come from and will happily move on from OpenAI et al. to DeepSeek-powered tech solutions. I hope Silicon Valley investors like Marc Andreesen end up taking a financial beating in this, but I don’t think DeepSeek will halt the encroachment of AI counterfeiting into our lives.
If anything, a better and more efficient system for generating fake answers, fake pictures, fake writing, and fake people will result in more of that stuff, not less.
Yet here I must make a confession of optimism and admit I don’t think the plan outlined above is going to work. The AI bubble is inflated in large measure thanks to its promise to help prop up platform bubbles across technology. (It also promises to help companies lay off and reduce the value of skilled employees, but that’s a different threat.) I think it might do that in the short term but I think in the long term it’s poison, because once people begin to suspect that nothing on these platforms is trustworthy, not even the people commenting and posting, you’re going to see a lot of users disappear and a complete failure to capture young audiences.
Bubbles are about the perception of value versus actual value. The history of tech platforms is one where consistent failures to generate actual value is masked by speculative values and acquisition valuations. That works when the whole ballgame is about attracting investors in a world where capital flows freely between markets and you can recoup massive losses with one successful sale (usually to another overcapitalized underperforming firm). But actual human attention is more fickle. The most likely outcome in your daughter’s case is not that she is just cluelessly guzzling fake movie trailers and watching fake influencers. It’s that she stops engaging with that type of content entirely because it all loses value to her. Even a lot of the real stuff, unfortunately, because the rampant counterfeiting is going to devalue it, and eventually she’s going to be looking for places where she doesn’t have to sift through quite so much bullshit to find something authentic. It won’t be a moral stand, it’ll just be a practical one. But all this shit is predicated on connecting people with stuff to sell with people who might potentially want to buy it, and once the sellers realize they’re paying to access phantom audiences, platforms that embraced AI slurry and bullshit growth metrics are going to be in trouble.
They’re already nervous about some of these trends I think, which is why they were so bullish on AI (it also made them look dynamic and innovative), and it’s the nature of a bubble that a lot of interested parties have no interest in popping it. Marketers who have relied on platforms for decades certainly don’t want to tell their clients they’ve wasted 95% of their spend accidentally promoting ads into the feeds of pornbots and disinfo accounts. But the platforms and middle men can only keep this going so far when you’re training these unprecedentedly large audiences that the global commons has turned into something less convincing than the floor show at Chuck E. Cheese.
Patrick: I think your instincts are right. Where I see AI slop show up for my oldest is in YouTube Shorts, their equivalent of shortform TikTok-like videos. They are often about showing something heroic, sad, or something in-between. It really is no different than the shrimp Jesus stuff that’s all over platforms like Facebook, engaged by a collection of bots and old people.
But young kids and old people have something in common: a lack of perception. Or they don’t care. The result is the same, though, a steady stream of thoughtless engagement with colors.
With my kid, at least, her reaction to these platforms, despite the slop, is to create. She recently filmed a “YouTube video” with her younger sister, coaching her on what to say, how to walk people through coloring a drawing, and how to “like and subscribe at the end.” The impulse was human. I take solace that beyond the misleading algorithms is a self-driven impulse for art, and you try to unpack that indescribable impulse using the platforms available when you’re alive.
These companies can try to bury it, but history shows us it will not be denied.
I mean, sure, she’s emulating a YouTube video, but that’s all you can do at eight years old, you know? You emulate what you know. You emulate, before you can wander into your identity.
The natural impulse of some parents, I think, is they’ve gotta teach their kids how to suss this bullshit out. I’m not so sure. I think kids are smart, and they will suss out the bullshit on their own. The rise of places like YouTube and Twitch, for all their many faults, is driven by a sense of authenticity—it’s a direct reaction to a long line of media institutions that sanded down personality into bylines. What we’re left with, of course, is often all personality and little substance, but I think the broader point stands. I do not think they will settle for slop. The slop might surround them, might be forced upon them, and they might (to our own frustrations) leverage tools of slop in creating art, but truly, I think they come out the other side mostly okay.
I am an optimist by nature, though. That’s in my bones. Why else have children in such a despairing age, you know? Because I have some faith they fight to emerge on the other side.