🔗 This is what ChatGPT is actually for:

But I’ve realized that feeling, of wanting to tell it more so that it can tell you more, is the multi-billion-dollar business that these companies know they’re building. It’s not fascist anime art or excel spreadsheet automation, it’s preying on the lonely and vulnerable for a monthly fee. It’s about solving the final problem of the ad-supported social media age, building up the last wall of the walled garden. How do you get people to pay your company directly to socialize online? And the answer is, of course, to give them a tirelessly friendly voice on the other side of the screen that can tell them how great they are.

On a recent episode of the Tech Won’t Save Us podcast, host Paris Marx was talking with 404 Media’s Jason Koebler and a topic that came up was a recent statement by Mark Zuckerberg about how he thought that the Facebook timeline would soon be just AI bots. The gist of the conversation was “This seems crazy, why would anyone want that?”

On the one, I agree—that does sound awful. I don’t want that, and it is hard to imagine how any other real person would want that either. Weirdly, though, I can also imagine that if Meta rolls this sort of functionality out, there will, in fact, be millions of people spending time on the platform, merrily interacting with AI bots.

I feel like a broken record on this topic these days, but everyone continues to talk about how Facebook—and the rest of these big platforms—isn’t giving users what they want, when we really need to be thinking about these product and architecture decisions in terms of what is good for the platforms. They are not in business of keeping giving their users what they want and keeping them happy; they are in the business of keeping their users engaged and clicking on their site so that they can sell advertising and make their stock price go up.

What has worked for Meta née Facebook so far to keep users engaged is to have a feed filled with content to engage with—to like, to dislike, to comment reply to and share. This approach has worked up to this point platform gets this content for free from it users and from publications and media companies that they have convinced to be on the platform. Sure, they have cloaked it all in marketing language about “creating connection,” but connection has never been what they are selling; connection (or the promise of it) has been the bait they have been offering to get us in their and creating content for them for free. And that content is not what they are selling either. It’s just more bait.

Nonetheless, I can imagine that even this roundabout means of attracting and keeping people stuck in their platform is highly annoying for Meta. They have to pick and choose from the content that is uploaded to their platform, and they have to deal with moderation because if you open up publication to everyone on the internet, some subset of that content is going to be The Worst Stuff On Earth. They also have to manage and moderate all of the relationships and interactions between their users.

How much better for Meta, then, if they can populate users’ feeds entirely with synthetic content that is mass-produced in house and is infinitely customizable. They take all the stuff they’ve learned about you from everything you’ve done on and off their platform, and feed that into their LLM systems to keep churning out an endless slurry of engagement bait.

I look at that prospect and think “Christ, what a load of shit,” but for Meta it’s a win.

Originally published at https://petebrown.bearblog.dev/facebook-has-never-been-about-connection-the-connection-has-just-been-bait/.