For a long time, big companies had a wall around them made of money. Enterprise software. Data teams. Round-the-clock support. Real personalization. All of it took a budget only the big players had. AI is knocking that wall down. What used to cost millions now runs a few hundred dollars a month, and a five-person shop can do things that used to take fifty people.

If you're small, this is the best opening you'll get. If you're big, the advantage you've leaned on — "they can't afford what we do" — is draining away right now. In three to five years every company will have AI and roughly the same efficiency gains from it. When everyone has it, it stops being an advantage. It's just the entry fee.

Hashi's take: The only question every business owner should be asking is: what makes us impossible to copy once everyone has AI?

If someone on your leadership team is still measuring AI by efficiency alone, forward it their way.

If you haven't figured it out yet, AI is the great equalizer. And that should give you all something to think about. Some of you will be thrilled by that. Some of you should be wary.

Size used to be the whole advantage. If you were big, you could throw money at things to keep your customers around — the analytics team, the personalization engine, the support desk that never sleeps, the marketing department. Smaller players couldn't match it. Not because they weren't good. Because they didn't have as much access to the resources needed.

That's changing fast. Tools that used to need an enterprise budget now cost less per month than a software subscription. A ten-person company can analyze customer behavior, forecast demand, personalize its marketing, and flag which customers are about to walk — work that used to require a dedicated data team — for a few hundred dollars a month, without hiring a developer. The kind of personalization that belonged to companies with deep pockets is now something a small shop can just switch on.

If you run a small business, this is your moment. You can offer the experience that used to be a large company's entire edge, and you can move faster doing it. No committee. No eighteen-month rollout.

If you run a large business, the competitor you didn't take seriously is going to be knocking on your customers' doors, matching your capabilities at a fraction of your overhead and cost.

Big companies have never needed to be more vigilant. Small companies have never had more to gain.

Most executive teams right now are asking the same handful of questions. Are we doing enough with AI? Do we have enough of it? What's our competition doing? Those are fair questions, and if you're not adopting, you will get left behind. But adoption isn't the finish line. It's the starting gun.

Let's be precise about what's actually new here. AI has been around for decades. What changed is agentic AI, built on top of generative AI — systems that don't just crunch your numbers, they do the work. And that's going from something you choose to something foundational. Gartner projects 40% of enterprise applications will have AI agents built in by the end of this year, up from under 5% in 2025. Your CRM, your accounting software, your scheduling tool — it's coming baked in whether you planned for it or not. There's no advantage in something everyone gets by default.

Think about what that actually means for you. Your competitor gets the same productivity bump, from the same tools, on the same timeline. You both get faster. You both get cheaper. And the gap between you doesn't move an inch — you just both end up with leaner teams and better margins that everyone in your industry now shares.

Efficiency everyone has isn't an edge. It's the new zero.

Which is why "are we using AI?" is the wrong thing to be measuring. Within a few years that question has the same answer for everyone in your market. Yes. And a question everyone answers the same way can't be the thing that sets you apart.

The numbers already show the split. Two-thirds of organizations report real productivity and efficiency gains from AI. But PwC surveyed more than 4,400 CEOs this year and only 12% had seen both revenue growth and cost reduction. Almost everybody is getting more efficient. Almost nobody is turning that into a business worth more.

Warren Buffett has a word for what's missing. He calls it an economic moat — the durable advantage that keeps competitors from chipping away at your profits. His test is simple. If a rival can copy it, it was never a moat.

AI efficiency fails that test. Anything your competitor can buy off the shelf next quarter isn't a moat. It's a subscription.

Take the efficiency gains. Bank them. But if that's where your AI strategy stops, you've built nothing that lasts.

Once the technology levels out, advantage moves to what AI can't hand your competitor — the data you own, the brand you've built, the relationships you hold, your own judgment. Those take years to build and none of them come in a subscription. Three places to start.

1. Inventory your data. Not a strategy deck — a list. Every system holding customer history, transactions, service records, and the notes your people keep in spreadsheets nobody else can see. Who owns each one, what shape it's in, whether anything could actually use it today. This matters because when everyone runs the same models, what you own counts for more than which model you use — more than half of investors now call proprietary data the strongest durable advantage a business can have. You can't build a moat out of data you can't find.

2. Rebuild one workflow from scratch. One. Not a company-wide transformation. Take the process that costs you the most or annoys customers the most, and redesign it around what's possible now instead of bolting AI onto the old steps. Bolting AI onto a broken process just gets you a faster broken process. The research is blunt: "layoffs are not a viable AI strategy." The companies pulling ahead are redesigning how the work happens, not cutting the people who used to do it.

3. Name what stays human — and defend it. Write down the moments where a real person is the product. The judgment call. The hard conversation. The relationship that took a decade to earn. AI writes your competitor's marketing copy in seconds, but it can't manufacture forty years of people trusting your name. Buffett bought See's Candies in 1972 and still owns it, because a brand people trust lets you charge more and keep customers longer. Rule those moments out of automation while your competitors race to strip theirs out.

Then make "what can't be copied here?" a standing question in your quarterly review. When AI is a given, that question is your strategy.

For the last three years the message has been simple: adopt AI or get left behind. That was right. It's also nearly finished. The adoption wave is going to crest, and when it does everyone will be standing on the same ground, roughly equally efficient, wondering why it didn't feel like winning.

The advantage goes to whoever uses this window — while everyone else is still congratulating themselves for finally using AI — to build something a competitor can't copy.

Buffett built an empire on one idea: find the castle with the widest moat. Tools change every few years. That question never does.

- Hashi