As a kid in the ’80s, I was captivated by Star Trek, especially the idea of traveling faster than light. It was pure fantasy, but it made me believe that with the right technology, even the laws of physics could be rewritten. I studied computer science and bioinformatics, hoping to help build the engines of that future.

Fast forward to today, we’re living in that future and then some. AI didn’t arrive quietly. It exploded, upending industries, rewriting playbooks, and erasing moats built over decades. We’re not just accelerating; we’ve hit terminal velocity. Here’s the paradox: The more powerful AI becomes, the more valuable our creativity, empathy, and human connection become. Being human is the edge.

From linear to ludicrous speed

In 2018, I believed the convergence of sensors, compute, and robotics would compress decades of progress into years. I was too conservative. Acceleration is now exponential. AI isn’t just analyzing data, it’s seeing what we can’t. Context windows have grown 100× in two years, enabling models to process entire codebases or simulate workflows in seconds. The cost of running massive models has dropped 1000×, putting what was once lab-grade tech into the hands of solo developers.

Then came agentic AI, digital experts that don’t just respond to prompts but research, reason, and execute entire projects autonomously. When top engineers told me they’d stopped coding because their AI agents were that good, I thought they were joking. They weren’t.

And this is still just the beginning. Google’s latest video model, VEO3, generates visuals so realistic they’re indistinguishable from reality. We’re not inching forward, we’re leapfrogging industries. Every time you find stable ground, it shifts.

Scarcity in the land of plenty

Everyone loves to talk about how AI drives abundance. But paradoxically, the most powerful capabilities are already becoming scarce. Compute, energy, and networking—the infrastructure AI needs—aren’t scaling fast enough to match demand. Running a network of AI agents isn’t like running traditional software. It is orders of magnitude more expensive.

We’re already seeing it. People who once balked at paying $10 per month for software now spend $200 or more on premium AI tools. Those prices are still subsidized by growth-stage ambitions. As companies shift to profitability and raise prices, many users will keep paying. Why? Because the value is real.

This means intelligence itself is becoming stratified. In the next three to five years, the best AI won’t be universal. It will be a luxury. That changes the entire economy of knowledge.

The art vs. science flip

As AI consumes technical work, human judgment becomes the last real moat. For decades, technical skill was the holy grail. Now, AI can replicate much of it instantly.

But AI doesn’t know what problem to solve. It can’t predict what will resonate. It can’t read a room. That’s the domain of humans: the executive who senses when to pivot, the founder who intuits product-market fit before the metrics appear, the storyteller who moves a room, the leader who earns trust in a single conversation. These aren’t lines of code. They’re moments of human clarity.

This shift flips the value stack. Skills that once earned the highest salaries get commoditized. What’s irreplaceable is judgment, the kind earned from thousands of ambiguous decisions. Those who lean into their experience and intuition become the real power players. Everyone else risks becoming middleware, human interfaces between AI engines, unless they rediscover their edge.

The speed imperative

In this new landscape, speed is the only sustainable advantage. AI capabilities evolve monthly. Market assumptions break quarterly. Disruption now comes from two-person teams using bleeding-edge stacks to ship faster than any incumbent can react.

I’ve seen it happen. Developer tools like GitHub Copilot, launched with huge fanfare, were quickly overtaken by startups like Augment Code. They moved faster, broke with traditional coding paradigms, and built natively for this new era. By the time others reacted, the market had already moved on.

Speed isn’t a bonus anymore. It’s survival. The companies that build for velocity and can make big decisions in days, not quarters, will be the ones that last.

The human edge in the age of AI

This isn’t just another tech cycle. Cloud, mobile, crypto—those changed how we access information. The AI revolution changes how we decide and act. It’s reshaping economic power around access to compute, and ironically, elevating the value of uniquely human insight.

For leaders, that means prioritizing creativity, intuition, and human connection over rote technical skills. For investors, it means betting on founders who can move fast and have good intuition. For everyone else, it’s a wake-up call: The machines can code, write, and reason. But they can’t be human.

That’s not a flaw. It’s what makes us indispensable.

The opinions expressed in Fortune.com commentary pieces are solely the views of their authors and do not necessarily reflect the opinions and beliefs of Fortune.

Read more: