The New Leverage: What AI Makes Possible That Nobody Has Priced In Yet

Insight

Oracle laid off 16,000 people. Block cut 40% of its workforce — engineers, PMs, entire teams. Meta grew revenue 700% while shrinking headcount. Marketing agencies are struggling to justify their fees. SEO agencies are losing clients who can't explain why they're still paying.

These aren't isolated decisions. They are companies that have looked at what AI can now do and restructured accordingly. They are the canary in the coal mine.

The storm isn't coming. It's already making landfall. The question is whether you understand the new economics well enough to move with it — or whether you're still operating on assumptions that are quietly becoming obsolete.

The asymmetry is real and it is large. What AI makes possible today is vastly underpriced relative to its output. The people and companies who understand this are compounding that advantage every day. The ones who don't are working harder for results that keep shrinking.

The Deep Dive

Two Kinds of Asymmetry

The first asymmetry is compression. AI lets a small team do what previously required a large one. Claude Code and Codex — two of the most sophisticated developer tools ever built — were each developed in days, not months, using AI to write, test, and deploy the code. A solo founder with clarity and the right tools now ships what used to require fifteen engineers and a quarter of roadmap.

I've been living this directly. I am not a developer. I don't write code. And yet I have built and shipped production AI systems — multi-agent architectures, full product experiences, content pipelines — that would have required an engineering team a few years ago. The leverage is not theoretical. It is real and it is available right now.

The second asymmetry is capability. AI lets you do things that were previously impossible — not just faster or cheaper, but structurally out of reach. Read every Reddit post in a niche and synthesize genuinely personalized customer profiles. Run a dark factory — an application that builds, tests, maintains, and improves itself autonomously while you provide direction through markdown PRDs. These are not future possibilities. They exist today, used by a small number of people who understand what they're sitting on.

This Is the 2008 Moment

Early Google Ads buyers paid pennies for clicks that now cost dollars. Early SEO practitioners built domain authority before anyone understood its value. Early Twitter builders grew audiences before brands arrived and made the game expensive.

In each case the window wasn't about being first to the technology. It was about being first to understand the new economics the technology created — and moving before the market priced it in.

That is where we are right now. The platform exists. The behavior is shifting. The new economics are visible if you know where to look. And the vast majority of companies and individuals are still operating on the old model, which means the asymmetry between what's possible and what's understood is at its widest point.

It will not stay this wide for long.

The next article is about what the new economy actually rewards — and how value gets created, surfaced, and captured when AI is the infrastructure everything runs on.

In this series

• 1. We Are Living Through the Biggest Shift Since the Internet

• 2. These are the Last Days of the Attention Economy

• 3. The Conversation Is the Work

• 4. The New Leverage: What AI Makes Possible That Nobody Has Priced In Yet

• 5. Forget Followers. Here Is How Value Actually Surfaces in the Age of AI.

• 6. The Imagination Age

• 7. Skills in the Age of AI: Execution Is Free. Distribution Is Handled. So what’s left?

Talk soon,

Stefan