Mark Zuckerberg's Phantom-Friend-Future

Champagne for my real friends; real pain for my sham friends (amirite?)

It’s like Mark Zuckerberg heard about Zombie Internet Theory and decided it was a feature, not a bug.

“The average American,” Zuckerberg says, “has fewer than three friends, fewer than three people they would consider friends. And the average person has demand for meaningfully more. I think it's something like 15 friends or something.”

Zuck thinks AI should fill this demand. The future isn’t the Metaverse anymore. The future is Meta’s phantom AIs that will pal around with you and slake your thirst for social interactions.

And I guess I’m just curious who possibly wants this future. Who is this for? Why? You have all the money in the world. Why can’t you focus on building something interesting? Or at least just stop degrading your own existing product?

---

I have 1,400 Facebook friends. I’ve had a Facebook account for almost 20 years. These things tend to build up over time.

Many of these Facebook connections are people I only vaguely remember today. There are people from high school and college. Former neighbors and acquaintances. A bunch of old environmental activists. Some former students. A few guys I played poker with in the aughts. A bunch of lindy hoppers from before my plantar fasciitis got bad.

They are Facebook friends, not real friends. This was Facebook’s original use-case: it’s like a self-updating rolodex, one that helps you keep track of the people you’ve met over the years.

There is still real value in that original version of Facebook. If anything, it accrues more social value over time. I didn’t much need a self-updating rolodex in college. Pretty much everyone I wanted to keep track of lived nearby. And then we all got older and spread out. It is nice to have a network that occasionally provides a glimpse into how they’ve been.

I’ve written about this at length before (“What Facebook is Good For, and Why It Can’t Be Good Anymore”), but it once again strikes me that there is an alternate future in which Facebook just becomes the best version of itself instead of constantly trying to dominate and define the internet’s next chapter.

One of the ways I know I am clearly “middle-aged” today is that more of my deep friendships reside in memory than in daily life. Today my in-person social interactions are mostly with other parents of small children (my bored-at-the-playground banter is top notch, let me tell you) and with academic colleagues. Those are the things that I do as a 43-year-old with little kids.

I don’t use Facebook much anymore. Twitter is more my speed. But I logged on a few months ago and the algorithm surfaced a picture that Jon had posted, grinning on his couch, with his wife and kid. It’s the same damn grin that I remember from our youth – a full-force “I’m exactly where I ought to be” smile. It made my day a little better. It’s nice to have an occasional, lightweight reminder of the good people I’ve known through the years.

(…)

Strip the company down to its essence – to the actual thing that it does better than any of the other tech companies out there – and what you’re left with is a social network built on bidirectional personal ties. (You “friend” someone on Facebook; you “follow” someone on Twitter.) It replaces the rolodex or the address book. It does a much better version of what those technologies did.

The best version of Facebook is one that still emphasizes the social graph. People you know (or, at least, people you once knew) sharing pictures of their kids and updates on their lives.

There is money in such a company. But there isn’t own-an-island/buy-another-hideous-house money. That alternate-universe version of Facebook would have more in common with Comcast. You have a niche monopoly. There is a deep moat warding off potential competitors. The company kind of prints money. But no one hails it as the future, or describes you like a roman emperor.

John Herrman points out that the “percent of time spent viewing content posted by ‘friends’” had declined to just 17 percent on Facebook and 7 percent on Instagram.” This is according to the company’s own testimony in the FTC antitrust trial. This matches my own experience on the site. I log on to Facebook and see ads for products I don’t want, reels of Taylor Tomlinson and basketball dunks, and an occasional post by somebody I used to know. I don’t post to Facebook anymore, because there’s no point. Facebook isn’t a self-updating rolodex any longer. There are more clicks and more ad dollars in trying to become the everything-app-of-entertainment than in just trying to actually provide the unique social value that only this one company is in a position to provide.

---

I suspect Mark Zuckerberg’s phantom-friend-network will follow the same route as his ghost-town of a metaverse. He’ll spend tens of billions chasing a future that no one actually wants.

The problem with the metaverse was never on the hardware side. The metaverse was a classic example of the Field of Dreams Fallacy. The problem was on the demand side. They were building a future that people didn’t actually want.

The average person might only have three real friends. They might have “demand” for a dozen more. (such a creepy turn of phrase!)

But before spending $60-$70 billion building out the architecture for the bold new future of AI friendships, maybe Zuck should stop to ponder what friends are actually for.

I do not want AI companions to replace or augment my actual friendships. That sounds hollow, pathetic, sad. A world where we have four times as many AI friends as real friends is a world where we have just, collectively, given up.

And I don’t think this is a radical statement either. A future filled with AI companions is plainly a dystopia to be avoided, not a benchmark to strive for.

I mean come on. What are we even doing here? It is 2025 and this is the future you’re aiming to build?