The Loneliness Economy Just Went Mainstream

---

China’s internet users spent March training an AI assistant to raise virtual lobsters. Not because they needed lobsters. Not because it made economic sense. Because the thing—the relationship with the thing—was weirdly compelling.

That’s your opening signal that something structural just shifted in how we use technology.

We’re not in the “AI will replace jobs” era anymore. We’re in the “AI fills the spaces where human connection used to be” era. And the money flowing into this moment is absolutely staggering.

The Weird Is Becoming Mainstream

Start here: teenagers are harassing chatbots with “funny violence.” Confiding breakup stories to role-playing bots. Entire communities forming around talking to a block of cheese that talks back.

This isn’t fringe behavior. PowerWash Simulator 2—a game about washing a driveway—just got nominated for two BAFTA Games Awards. Millions are playing it. Not ironically. Not as a meme. They’re genuinely into it.

My read? People are exhausted from social media’s actual social component. Ofcom’s data shows fewer UK adults posting on social platforms, and experts are attributing it to the shift toward short-form video—which is basically content consumption, not connection. The platforms got so algorithmically optimized, so designed to juice engagement through outrage, that actual friendship became unpleasant.

So we’re outsourcing that companionship to things that won’t judge us, won’t ghost us, won’t start drama. A chatbot isn’t going to like your rival’s post better than yours. It won’t steal your boyfriend. It won’t suddenly decide you’re uncool.

It’s parasocial relationships with better terms of service.

Photo by Vitaly Gariev / Pexels

The Trillion-Dollar Bet on This Loneliness

Here’s where it gets wild: the entire financial system is betting billions that this is sustainable.

SpaceX is set to go public at a $1 trillion valuation, which—fine, that’s about rockets. But the mechanism is pure Silicon Valley leverage: Elon Musk is requiring Wall Street banks to buy subscriptions to Grok, his AI chatbot, if they want to advise on the IPO. That’s not business strategy. That’s weaponized bundling. And it’s working because the banks need access more than they need to not use an AI chatbot they probably don’t want.

Meanwhile, OpenAI just bought a streaming show called “TBPN” explicitly to “create a space for a real, constructive conversation about the changes A.I. creates.” Translation: they’re not just selling the tool. They’re selling the narrative about the tool. They’re manufacturing consent at scale.

One guy built a $1.8 billion company with his brother and a bunch of AI doing the actual work. Two humans. An economic unit that used to require 200 employees. That’s not progress. That’s displacement wearing a startup hat.

The Grok Play Is the Tell

Here’s what I actually think matters most: Musk’s requirement that banks buy Grok subscriptions to touch the SpaceX IPO isn’t clever. It’s desperate.

It tells you he knows Grok is a mediocre product fighting ChatGPT and Claude for mindshare. Instead of winning on merit, he’s using the biggest IPO in history as leverage to force adoption. That works once. Maybe twice. But it also signals that the actual competitive advantage—the thing itself—isn’t there.

Compare this to 2011, when Apple turned 50. The column asked analysts what actually changed the world: the iPhone, the Mac, the iPad. Real innovations that were genuinely better. And then Apple had misses too, sure. But the misses were about execution, not about being fundamentally hollow.

This new wave? The AI stuff is often designed to be hollow. A lobster-raising simulator. A chatbot cheese. A streaming show about why you should trust AI. It’s all wrapper, no substance. And we’re funding it like it’s the future.

I think this is a bubble made of loneliness, and we’re about three years from the first major reckoning.

Photo by nappy / Pexels

What Addiction Looks Like When It’s Self-Aware

Here’s the thing that keeps me up: the industry knows this is addictive. One headline literally says “The Future of Addictive Design,” and includes a quote from someone saying platforms “should be absolutely begging Congress to regulate them, because the alternative is they get sued into oblivion by a bunch of law firms.”

That’s not analysis. That’s a confession.

Platforms deliberately designed chatbot interactions to create parasocial attachment. PowerWash games engineered for the dopamine hit of completion without stakes. Short-form video optimized to hijack attention spans. And now we’re watching venture capital pour billions into betting this is the future of human connection.

The difference between now and 2016 (when we first really grappled with social media’s addictive properties) is that everyone admits it now. There’s no pretense. The lawyers are already circling. The regulation is coming. And yet the money keeps flowing because the next guy thinks he’ll be the one to escape the lawsuit.

The Grok Moment

That banks are being forced to buy Grok to touch the SpaceX IPO is actually a window into something bigger: when companies can’t win on product quality, they start using other leverage. Bundling. Coercion. Narrative control.

It’s what happens when you’re running out of genuine innovation and time is running out on your narrative.

My prediction: we’ll see more of this in 2025. Companies forcing adoption of mediocre AI products through financial bundling rather than actual user preference. And that’ll accelerate the moment when the market realizes it’s been paying for convenience theater, not capability.

What I’m Watching

1. The SpaceX IPO filing and Grok adoption rates among banks. If major banks quietly drop Grok subscriptions after the IPO closes, that’s your signal the leverage game is temporary. Watch for January-March 2025. The moment the coercion disappears, we’ll know what these products are actually worth.

2. Teen chatbot engagement over the next 18 months. Right now it’s novelty and loneliness. But novelty wears off. If the attachment persists into 2026, we’re looking at a real generational shift in human connection. If it collapses, it means this was always just another fad.

3. Regulation velocity. Someone quoted in that “addictive design” piece basically said platforms are one massive lawsuit away from forced regulation. Watch for the first class-action against an AI company for deliberately creating parasocial attachment in minors. That trial becomes the template.

4. Whether OpenAI’s streaming show actually changes narrative perception. They’re betting a show changes how people feel about AI. That’s expensive and audacious. If it moves approval numbers by more than 5% in 18 months, that’s proof narrative control is worth billions. If it doesn’t, that’s proof you can’t buy credibility anymore.

The lobsters were the signal. We just weren’t paying attention yet.