Silicon Valley's Weirdest Year: When Billionaires Play Hardball and Nobody Wants to Post

---

The Grok Tax on Wall Street

Let me start with the wildest thing happening right now: Elon Musk is literally gatekeeping SpaceX’s IPO behind a Grok subscription paywall.

This isn’t normal billionaire theater. Banks that want to advise on one of the largest public offerings in history—SpaceX valued at $1 trillion in its planned listing—have to buy subscriptions to Musk’s AI chatbot. It’s the financial equivalent of a bouncer saying, “You can enter, but first you gotta buy a drink at my bar.”

What’s remarkable isn’t the arrogance. It’s that Musk’s doing this while simultaneously positioning SpaceX for a historic valuation. The IPO could make him the world’s first trillionaire. He’s not trying to hide his leverage. He’s weaponizing it.

Photo by Griffin Wooldridge / Pexels

This tells you something crucial about how power flows in 2025. Traditional gatekeepers—bankers, institutional investors, regulators—don’t have the asymmetrical control they used to. Musk can afford to offend them because his company’s worth justifies the insult. They need him more than he needs their goodwill.

But here’s my read: this move reeks of someone who’s testing boundaries because he’s gotten drunk on winning. SpaceX doesn’t need the Grok subscription to happen. The money’s already committed. This is pure flex, and it’s going to poison some relationships when it matters.

Meanwhile, Nobody’s Posting Anymore

Contrast that with something quietly seismic: UK adults are posting less on social media. Ofcom found it. Experts link it to platforms chasing short video—TikTok, YouTube Shorts, Instagram Reels—which are fundamentally passive consumption engines, not creation tools.

This is the inverse of what happened in 2009-2014, when everyone wanted a feed to broadcast their life. Facebook, Twitter, Instagram—they all thrived because posting felt like being. Now the algorithm rewards watching, not sharing.

The shift’s already happened. You can see it in behavior. People still scroll endlessly. They just don’t upload selfies or hot takes anymore. They’ve become viewers instead of performers.

I think this collapse in posting will be remembered as the moment social media stopped being about connection and started being about extraction. Platforms get your attention. You get the anxiety of comparison without the dopamine of an audience.

The AI Chatbot Gold Rush Has Friction

Here’s where it gets chaotic: AI companies are discovering they can’t actually scale their flagship products without running into hard limits.

Anthropic’s Claude Code users are hitting usage caps “way faster than expected.” OpenAI just bought a streaming show (TBPN) to improve its narrative around AI. And someone built a $1.8 billion company with two employees and a lot of AI automation—which the headline describes as “super efficient and a little bit lonely,” which is maybe the most honest thing written about AI this year.

These are three different problems masquerading as one. Anthropic’s hitting infrastructure walls. OpenAI’s worried about perception (they’re literally buying media to shape the conversation). And the bootstrapped AI company story is technically impressive but also kind of dystopian.

What’s missing from all this is someone admitting: we don’t actually know if people want these tools, or if they’re just the next shiny thing. The usage limits suggest demand outpaced supply. But demand for what? Actual work? Or novelty?

Photo by nappy / Pexels

The Crime Nobody’s Talking About: Hasbro Got Hacked

Peppa Pig and Transformers owner Hasbro suffered a cyber-attack. Operations stay open. “Some delays” possible.

No one’s talking about this. There’s no congressional hearing scheduled. No outrage. A company that owns the IP for children’s most-watched shows got compromised and we’re just… acknowledging it as a news item.

This is how we’ve normalized cybersecurity failures. It’s background radiation now.

Games About Washing Virtual Toilets Are Winning

PowerWash Simulator 2 got nominated for two BAFTA Games Awards. This is happening because millions of people would rather play games about mundane jobs—mowing, cleaning pools, power-washing—than experience actual labor.

There’s something almost poetic about this. We’ve automated so much that the fantasy is… doing a simple job well.

Real work often feels meaningless. Games about cleaning are entirely meaningful. You get real-time feedback. The toilet gets clean. You feel progress. Then the level ends and you move to the next one. No meetings. No Slack. No performance reviews.

I think this is revealing what people actually want from games: not escape, but efficacy. A world where effort produces visible results. Where you own the outcome.

The Regulatory Bomb That Everyone Sees Coming

Here’s my genuine uncertainty: I don’t know if social media platforms will get regulated before they self-destruct or if they’ll collapse first.

One expert quoted in the headlines says platforms “should be absolutely begging Congress to regulate them, because the alternative is they get sued into oblivion by a bunch of law firms.” This is the smartest thing anyone’s said about tech regulation. Voluntary regulation looks better than lawsuits. But none of these companies are voluntarily doing anything.

TikTok’s facing a potential ban. Meta’s fighting over data privacy. Twitter’s hemorrhaging advertisers. And nobody knows if regulation happens in 2025 or 2028.

What I know: the current approach (move fast, apologize rarely, hire better lawyers) has an expiration date. The only question is whether that date arrives before the next election or after.

What I’m Watching

SpaceX’s IPO trajectory and whether the Grok gambit actually affects bank participation. If major institutions just… ignore the subscription requirement and advise on the deal anyway, Musk’s leverage gets exposed as performative. If they actually comply, we’re in a new era where billionaires can extract ancillary revenue from their own market events.

Usage patterns on social platforms for the next two quarters. If posting keeps declining while watch-time stays flat, that confirms we’ve moved to a pure extraction model. If watch-time drops, platforms have a real problem—they’ve optimized for engagement so aggressively that they’ve killed the thing people actually want (connection with humans).

Anthropic’s response to Claude Code scaling issues and whether they implement hard caps or open new tiers. This is the real test of whether AI assistants are genuinely useful or novelty tools hitting their actual market ceiling. Hard caps = scarcity = premium pricing = luxury product. If that’s the model, AI isn’t transforming work. It’s becoming a high-end tool.

Whether any major platform actually accepts regulatory pressure or if we hit the lawsuit phase in 2025. Once the first big settlement lands, dominoes fall fast. Watch for the first major class-action win, not the policy announcement.