Will AI Shrink Big Business' Advantages?

AI looks to be lifting small teams more than enterprises.

When I was in college, I raced road bikes. I trained a lot, but I wasn’t, let’s call it, a naturally talented athlete. Some people are just born with it. So I spent a lot of time in and at the back of the main group of riders—called the pack, or the peloton.

One of the most painful things in bike racing is what happens when the peloton hits a series of sharp turns. The riders at the front have to brake into each corner, then surge out of it. That slow-fast move ripples backward, and by the time it reaches you (me) at the back, the adjustments have compounded into something brutal: a near stop, then a sprint to close a gap you didn’t create.

But all that braking up front slows the whole pack down. And when the leaders are forced to keep slowing, the riders behind get a chance they wouldn’t otherwise have to close the distance, even to catch up.

There’s a version of this dynamic playing out in the market right now, as big companies struggle to gain an advantage with AI.

IS THIS THE END OF THE FIRM?

In 1937, an economist named Ronald Coase wrote a paper called “The Nature of the Firm.” It’s a defining paper, because he essentially asked: if markets are so efficient, why do companies exist at all?

The basic understanding we’ve drawn from it is that companies exist because of coordination. Searching, negotiating, and contracting through the market is expensive—it’s cheaper to pull all of that in-house. If your organization is big enough to have a design group, an IT group, or an HR group, you can simply email them. That’s the premise on which organizations are built.

It is also a tax.

If you’ve ever worked inside an organization, you know. There’s a ton of cognitive overhead. That “run it past legal so HR can sign off” friction that we affectionately call bureaucracy.

AI is removing the inefficiencies in the external market. It’s enabling more entities to work together more simply and with greater coordination.

This New York Times piece about small business owners managing whole armies of AI employees is what got me thinking about this new dynamic. They interviewed a bunch of people for the article, but the main persona throughout is Scott Bell, a bankruptcy lawyer who has been using OpenClaw to do all sorts of things. He bought four Mac Minis to run five agents, each acting out different tasks. Reading court notices. Messaging clients in Spanish. Drafting client counterproposals.

The power of AI, when you’re an individual, converts straight into the work. No corporate coordination tax is slowing things down.

Every week, there are articles about another CEO or CFO saying, “Hey, we’re not getting the uptick from AI that we were hoping for.” If you’re working in a large organization, trying to do the same thing, you feel that friction and understand why AI isn’t working as well as it should.

When you’re using AI inside an organization, the tax — the organization’s whole reason for being — is what protects it from agents doing bad things, which they can do. But it’s also preventing the AI from doing some incredible things. My hypothesis is that’s why there isn’t that lift. But that friction isn’t there to slow solopreneurs or small teams.

CATCHING THE PACK

What we’re going to see is AI enabling not just people to start their own businesses, but also to compete at a higher weight class. It used to be that scale, the size of the organization, created the efficiency and the capacity to do work that smaller groups couldn’t. Now, with a tool like AI, smaller organizations can do more.

Back in the Peloton. Either I got dropped and never caught the pack after suffering that fast-slow whipsaw three, four, five times in a race. Or if I was lucky, there were a few tight corners in a row, slowing the pack. A modest bottleneck was to the chaser’s benefit. That’s what’s happening with large organizations. The people in the back—the small companies—can start catching up.

Coordination tax is the bottleneck, and small groups using AI can finally catch up, if not exceed, the enterprise. Large organizations that require extensive coordination and incur significant overhead are vulnerable. That’s why scale, the old advantage, is starting to look like a liability.

In February 2026, the global law firm Baker McKenzie announced it was cutting up to 1,000 roles—and cited its growing use of AI as one reason. The cuts weren’t lawyers. They were the support layer: research, marketing, know-how specialists, and secretarial staff. Bloomberg Law reported the reductions hit up to a tenth of the firm’s global business-services workforce. Clifford Chance and Irwin Mitchell had made similar moves months earlier. Consulting is feeling it too; even McKinsey has started trimming its support functions.

Now, plenty of people think the AI explanation is too convenient. As one legal recruiter put it to the New York Law Journal, blaming AI lets a firm cut hundreds of people without looking cruel. Maybe. But that critique concedes that the work those people did is now optional. Whether AI is the cause or just the cover, the coordination layer is the first thing to go.

Scott Bell, the lawyer in the article, is using OpenClaw bots to keep his business running while he takes longer lunches. Others are making real money with side hustles. And for those who are motivated to do more, you’re going to see a huge uptick in businesses punching above their weight class.

I’m excited to see how much entrepreneurial heft will be thrown at the market.

The whole reason some people are finding real lift with AI while companies bemoan their token costs is that the premise of what a firm is for no longer holds as tightly.

So here’s the question: will large firms eventually wring the same lift out of AI that small ones already are?

Maybe. The coordination tax that’s slowing them down today is also what’s kept them safe — from agents going rogue, from the chaos that small operators are absorbing in real time. But I wouldn’t bet against the upstarts. The premise that built the firm—that coordination is worth the cost—is precisely what AI is pulling apart. And the people at the back of the pack have never been faster.

P.S. I got third in my last race of senior year in college. My best finish ever. It took everything I had, but I was glad to finally finish in front.