I've been thinking about how AI will affect ad-supported "content" platforms like YouTube, Facebook, Twitter, porn sites, etc. My prediction is that as AI-generated content improves in quality, or at least believability, they will not prohibit AI-generated content, they will embrace it whole-heartedly. Maybe not at first. But definitely gradually and definitely eventually.

We know that these sites' growth and stability depends on attracting human eyeballs to their property and KEEPING them there. Today, that manifests as algorithms that analyze each person's individual behavior and level of engagement and uses that data to tweak that user's experience to keep them latched (some might say addicted, via dopamine) to their app on the user's device for as long as possible.

Dating sites have already had this down to a science for a long time. There, bots are just part of the business model and have been for two decades. It's really easy: you promise users that you will match them with real people, but instead show them only bots and ads. The bots are programmed to interact with the users realistically over the site and say/do everything short of actually letting two real people meet up. Because whenever a dating site successfully matches up real people, they lose customers.

I hope I'm wrong, but I feel that social content sites will head down the same path. The sites will determine that users who enjoy watching Reels of women in swimsuits jump on trampolines can simply generate as many as they need, and tweak the parameters of the generated video based on the user's (perceived) preferences: age, size, swimsuit color, height of bounce, etc. But will still provide JUST enough variety to keep the user from getting bored enough to go somewhere else.

It won't just be passive content that is generated, all those political flamewars and outrage threads (the meat and potatoes of social media) could VERY well ALREADY be LLM-generated for the sole purpose of inciting people to reply. Imagine happily scrolling along and then reading the most ill-informed, brain-dead comment you've ever seen. You know well enough that they're just an idiot and you'll never change their mind, but you feel driven to reply anyway, so that you can at LEAST point out to OTHERS that this line of thinking is dangerous, then maybe you can save a soul. Or whatever. So you click Reply but before you can type in your comment, you first have to watch a 13-second ad for a European car.

But of course the comment was never real, but you, the car, and your money definitely are.