The Great AI Sorting: Who Wins, Who Loses, Who Gets Left Behind

---

The Oscars just banned AI actors and writers from winning awards.

The Pentagon just signed eight new contracts to weaponize AI.

These two things happened in the same news cycle, and nobody seems to notice they’re describing the same moment from opposite angles: we’re drawing lines in the sand about what AI gets to do—and those lines are wildly inconsistent.

Photo by Krizjohn Rosales / Pexels

The Line Hollywood Just Drew

Let’s start with what’s clearest. The Academy, which controls the Oscars, issued new eligibility rules on Friday saying AI-generated actors and writing cannot win awards. This isn’t a ban on using AI in filmmaking. It’s a ban on letting AI claim the award itself.

This matters because it’s explicit. It’s not “we’ll evaluate case-by-case” or “it depends on the context.” It’s: no.

But here’s what’s interesting about drawing a line so publicly—you immediately reveal where you think the line should be. And the Oscars’ line is basically: AI is a tool, not a creator. Tools don’t get trophies. Only humans do.

That sounds good. It sounds like we’re protecting human artists. But it’s also a defensive move by an institution that’s already bleeding cultural relevance. The Oscars’ viewership has collapsed from 55 million in 1998 to roughly 19 million in 2023. They’re not banning AI because they’re principled. They’re banning it because they’re scared.

The Pentagon’s Opposite Calculation

Meanwhile, the Pentagon signed eight new contracts with big tech firms to expand its AI capabilities and is explicitly positioning the US military as an “AI-first” fighting force.

Think about the asymmetry here. The Pentagon is saying AI isn’t just a tool—it’s a strategic advantage. It’s not just a component of warfare. It’s the foundation. This comes alongside separate agreements with companies like Anthropic to expand classified work using AI systems.

The military doesn’t care whether AI gets to “claim” the victory. They care that it wins.

And that’s the actual story. We’re not having a conversation about whether AI should exist or what it should do. We’re having a conversation about who gets to use it and toward what end. The Oscars want to preserve human achievement in storytelling. The Pentagon wants to preserve American military dominance. Those aren’t the same thing at all.

Photo by nappy / Pexels

The Musk Paradox

Now throw Elon Musk into this, because of course he’s in it.

His $158 billion Tesla pay package—the largest compensation deal in corporate history—comes with a catch: he can’t actually pocket it until Tesla hits a series of enormously ambitious milestones. So far he hasn’t. The money’s trapped in a conditional agreement, which means Musk is simultaneously the richest guy on paper and the guy who can’t access his wealth.

It’s almost poetic. Musk has spent the last two years warning about the dangers of AI while simultaneously building AI systems through his companies. He’s fighting OpenAI in court over concerns that it’s too focused on profit and not focused enough on safety. But the trial judge just ruled that jurors probably won’t hear Musk’s main argument—that AI could pose an existential threat to humanity.

So Musk’s fears about AI are literally too big to litigate.

Meanwhile, SpaceX is already traded on secondary markets through special purpose vehicles, meaning ordinary investors are holding SpaceX stock before any IPO. The company’s treated as inevitable. The exit is inevitable. The money is inevitable.

But Musk himself? Trapped. Can’t touch it. Conditional.

What Spotify’s “Verified” Badges Actually Signal

Spotify just rolled out “Verified” badges to distinguish human artists from AI ones. They’ll check live dates, social media presence, and other signals of human authenticity.

This is where the market starts catching up to the philosophy. Spotify isn’t banning AI-generated music. It’s creating two tiers. AI stuff can exist, but it gets marked. It gets sorted.

Think about that word—“verified.” It’s doing real work. It’s not saying AI music is bad. It’s saying human music is certifiable. You can trust it. You know where it came from.

We’re in the certification phase of AI integration. Not rejection, not embrace—taxonomy.

The Phone Addiction Escape Hatch

Here’s where I’ll admit genuine uncertainty.

One of the headlines in the pile is about phone addiction remedies, and the experts quoted say counting minutes doesn’t work. Real solutions involve mindful parenting, curated content, human connection. You know what I notice about that? It requires intention. It requires someone to decide to do something differently.

And I don’t know if we’re capable of that anymore.

We’ve optimized the hell out of engagement. Spotify knows what song you’ll click next. TikTok knows what video you’ll watch next. Tesla knows how fast you’ll drive next. The entire architecture is built around prediction and personalization.

So when experts say “human connection” is the antidote to phone addiction, they’re describing something that requires you to resist the system you’re already in. That’s not a remedy. That’s a revolution. And I genuinely don’t know if we’re having one.

Photo by Markus Spiske / Pexels

Why Ask.com’s Death Matters More Than You Think

Ask.com shut down on May 1 after nearly 30 years. Jeeves went with it.

You probably haven’t used Ask.com since 2003. But it was there. It was a reminder that the internet wasn’t always Google. It wasn’t always algorithmic ranking. You could ask a question in natural language and someone (at first) or something (later) would try to answer it.

The internet Jeeves represented was slower and more human. You’d ask. You’d wait. You’d get an answer. It assumed you might not know the perfect search terms. It assumed conversation.

That internet is officially dead now.

And we replaced it with something faster that knows what you want before you ask, which sounds great until you realize it also knows what you’ll click on, what you’ll buy, what you’ll believe. The trade was speed for autonomy.

My Read

I think we’re sorting AI into categories, not deciding whether to use it.

The Oscars says: no AI in our awards ceremony. The Pentagon says: all in on AI for warfare. Spotify says: AI is fine, but label it. Musk says: it’s dangerous and also I’m building it and also I can’t cash my check.

These aren’t contradictions. They’re evidence of how fast this is happening and how little consensus exists about what it means.

Here’s what I’d bet on: the next two years will look a lot like the next two years of social media policy. We’ll see lawsuits, regulatory moves, corporate disclaimers, public hand-wringing, and basically zero actual change in how these systems operate. The Pentagon will expand AI in classified work whether Congress votes on it or not. Hollywood will figure out how to use AI while maintaining human-facing branding. Spotify will make “Verified” badges feel warm and trustworthy while promoting AI music that sounds human enough.

The real line isn’t between AI and human. It’s between who gets to decide the rules.

What I’m Watching

• Tesla’s milestone timeline. Musk can’t touch his $158 billion without hitting specific targets. Watch whether Tesla announces which milestones they’re tracking publicly by Q3 2024. If they don’t, assume he’s not close. If they do, assume they’re confident. This is money that moves only if he delivers.

• Pentagon classified AI scaling. The Department of Defense signed deals for expanded classified work with Anthropic and others. Monitor whether any whistleblowers or congressional briefings reveal what systems are actually being deployed in the next 18 months. Classified work moves slow—if they’re moving fast, that’s the signal that this is urgent.

• Spotify’s “Verified” adoption rates. Will artists pay for verification? Will listeners actually trust it? Watch the first earnings call where Spotify breaks down how many artists completed verification vs. how many started the process. Abandoned processes signal the badge is theater, not real.