The 100-Foot Wave Just Arrived

OpenAI released GPT-5.5 on Thursday. Six weeks after GPT-5.4. During the press briefing, Greg Brockman admitted, "there are enough model releases that it’s probably getting hard to distinguish one from another." The president of OpenAI just told you innovation is outrunning his own ability to narrate it.

Meanwhile, most enterprises are still running pilot programs they greenlit in 2024.

People love to talk about the “AI capability overhang,” the gap between what AI can do and what organizations are actually doing with it. I’m sick of that framing. It makes the problem sound like a distance you close by walking faster. It isn't. The gap isn't linear. It's categorical. The wave changed. The old technique doesn't work anymore. You don't close a capability overhang by doing what you've always done with a slightly bigger budget. You close it by learning an entirely different way to ride.

Last summer I argued in this column that you can buy AI, but you can’t fake capability, that the firms who win this cycle won't be the ones with the biggest licenses or the flashiest pilots, but the ones that built genuine organizational muscle around the technology. That argument hasn't weakened. It's intensified. Capability is the price of admission. Method is what separates the companies that ride the wave from the ones it lands on.

Big Wave Surfing Isn't Surfing

Here’s what Kotler got right that the business world hasn't absorbed: the progression in extreme sports didn't come from athletes getting incrementally better at existing techniques. It came from inventing new methods that made the old constraints irrelevant. Tow-in surfing didn't improve paddling. It replaced it. And once that paradigm shifted, records that had stood for decades fell in months.

The same dynamic is playing out in business right now. Bank of New York has been testing GPT-5.5 alongside frontier models from Anthropic and others. Its CIO, Leigh-Ann Russell, said the improvements in accuracy are creating a "step change" and that “the more quickly we can establish that accuracy, the faster we can scale our models to have completely redesigned financial services.”

Redesigned. Not optimized. Not enhanced. Redesigned. That’s not paddling harder. That's tow-in surfing.

3M — 122 years old, known for adhesives and abrasives — launched 64 new products in a single quarter last year, up 70% year over year. It’s on pace to launch over 1,000 new products by 2027, expanding into aerospace, defense and advanced infrastructure. It committed a third of its $3.5 billion R&D budget to AI-powered research and simulation tools. Its CTO said AI will "fundamentally change how we do research." A century-old industrial company is now exploring domains that would have required years of learning — because the cost of understanding new markets collapsed to near zero.

It didn’t get smarter. It changed the method.

The Organizational Flow State

Kotler's research showed that extreme athletes don't just tolerate risk — they've learned to trigger flow states reliably, because at the level they perform, flow isn't optional. It's survival. The 100-foot wave doesn't care about your credentials or your experience. It only cares whether you can process information and make decisions fast enough to stay alive.

AI doesn’t care about your legacy systems, your org chart or your change management framework either.

The companies that will thrive are the ones that develop what I’d call organizational flow — the ability to sense, absorb and act on new capabilities in weeks, not quarters. Not one pilot at a time. Ten explorations running in parallel. Research conducted, feasibility assessed, directional decisions made — simultaneously, continuously, at a rhythm that matches the pace of the technology.

Two years ago, that was constrained by headcount and budget. Today it’s constrained only by your ability to process what you're learning and decide what to do with it. The bottleneck moved from knowledge acquisition to organizational metabolism.

And here’s what most people miss: when you explore 10 domains instead of three, you don’t just find seven more opportunities. You find connections between them. An insight from materials science informs your packaging innovation. A regulatory shift in one market signals what’s coming in three others. A company exploring one domain gets information. A company exploring 10 gets pattern recognition. That's the kind of advantage that compounds and can't be replicated from a single lane.

You Need A Jet Ski

The waves are getting bigger. OpenAI will release another model in weeks. So will Anthropic. So will Google. Every major frontier lab is shipping capabilities faster than any procurement cycle, board approval process or strategic planning cadence was designed to handle.

You can keep paddling. You can run your annual strategic planning offsite, budget for two or three cautious explorations and tell yourself you’re being prudent. But the 100-foot waves are already here. And the companies that learned to ride them aren't waiting for you to catch up.

Laird Hamilton didn’t look at a 100-foot wave and say, "I need to paddle harder." He said, “I need a jet ski.”

What’s your jet ski?