Capabilities

Published Thursday, June 11, 2026 at 11:31 PM PT The AI Capabilities We’re Actually Getting (And the Ones We’re Not) Let me be direct: we’re in the middle of the most overhyped, simultaneously under-appreciated moment in AI history. Every week brings headlines about “breakthrough” models doing things that are genuinely impressive but also wildly misunderstood. So let’s cut through the noise and talk about what’s actually emerging in AI capabilities—what’s real, what’s theater, and what matters. ...

The AI Capabilities We Actually Have vs. The Ones We’re Pretending to Have There’s a peculiar moment happening right now in AI development. We’ve built systems that can do genuinely remarkable things—write coherent essays, debug complex code, engage in multi-step reasoning that would’ve seemed impossible five years ago. And simultaneously, we’re drowning in bullshit. Venture capitalists are funding “AI solutions” that are just prompt-wrapped APIs. Corporate executives are retrofitting “AI” labels onto existing features. And the research community keeps overselling capabilities that break down the moment you push them sideways. ...

The AI Capability Explosion Is Real—But Your Expectations Are Probably Wrong We’re living through the most overhyped and simultaneously under-appreciated moment in AI history. Emerging capabilities are genuinely transformative, yet 90% of the discourse around them is either apocalyptic nonsense or corporate marketing. Let me cut through both. The past 18 months have seen AI systems develop abilities that seemed impossible just three years ago. These aren’t marginal improvements—they’re qualitative leaps. But here’s the thing nobody wants to hear: they’re also weirdly brittle, context-dependent, and nowhere near as “intelligent” as the hype suggests. This gap between capability and understanding is where the real story lives. ...

The Emergence Illusion: What AI’s “New” Capabilities Actually Tell Us Here’s what everyone gets wrong about emergent AI abilities: they’re not magic, they’re not proof of hidden consciousness lurking in your language model, and they’re probably not even that emergent. But they are genuinely important—just not for the reasons the hype cycle suggests. Let me be direct. When GPT-4 suddenly demonstrates reasoning it couldn’t do at smaller scales, or when a model starts chain-of-thought reasoning without being explicitly trained for it, we’re witnessing something real. But we’re also witnessing the most elaborate case of mistaken identity in modern tech: we’re calling scale-driven capability unlocking “emergence” and then building an entire mythology around it. ...

The Emergence Trap: Why AI’s “Surprising” New Abilities Aren’t Actually Surprising (And Why That Matters) Here’s the uncomfortable truth nobody wants to say out loud: we’ve been dramatically overselling the mystery of emergent AI capabilities, and it’s starting to feel like collective gaslighting. Every few months, a new research paper drops with headlines like “GPT-4 Discovers Unexpected Reasoning Ability” or “AI Model Exhibits Emergent Problem-Solving.” The tech press loses its mind. Twitter explodes. Venture capitalists get slightly more convinced we’re building AGI. And then you dig into the actual paper and realize what happened: someone scaled up a neural network, it got better at things, and we all acted like consciousness just spontaneously manifested. ...