AI isn't a gold mine. It's a gold rush. And everyone has a pickaxe.

What HBO’s *Deadwood* Tells Us About Today’s AI World.

Starting now, I’ll be posting here weekly. Not about which AI tool you should try this week. Not about yet another “10 prompts that’ll cut your workday in half.” There are already plenty of people covering that.

I’d rather ask one question: How do you turn the complexity of the AI market into meaningful growth in your B2B commercial processes—without getting lost in the tools? This first article starts one step earlier: a diagnosis of the world of AI.

Fortunately, at home I’m not always preoccupied with technology. Sometimes I just watch a show with my girlfriend. Right now, we’re watching* * *Deadwood*, the 2004 HBO series about a lawless gold rush town in South Dakota. No laws, no government—just people who want to get rich as fast as possible. What makes the series so interesting isn’t the gold itself, but the behavior of the people around it.

You don’t have to be familiar with the series to understand the parallel.

Who actually gets rich during a gold rush?

Deadwood isn’t really about gold. It’s about what gold does to people.

Everyone talks about gold. Everyone sells the tools. Everyone knows someone who supposedly knows exactly where to dig. But the real winners are probably not the people shouting the loudest.

As soon as a rumor spreads that there’s wealth to be had, everyone springs into action. Gold prospectors. Merchants. Opportunists. Powerful players waiting for the chaos to grow big enough to take over everything.

Everyone thinks they have a claim to a golden future. Everyone buys a pickaxe. Everyone points to claims where, according to someone else, gold is supposed to be buried.

But in a gold rush, not everyone gets rich. In a gold rush, the people who make the most money are often those who never find a single gram of gold themselves. Through the saloon where all the stories converge. Through the bank that decides who can scale up. Through the railroad that provides access to the market. Through the suppliers of shovels, sieves, and pickaxes.

And, of course, by the few who actually have gold on their claims.

But that last person doesn’t shout it from the rooftops. Anyone who truly has a rich claim doesn’t want prying eyes. They don’t put up a sign by the roadside saying, “The gold is here.”

Back to the present

I think that’s exactly what’s happening in AI right now.

The saloon is full of stories. On LinkedIn, X, in newsletters, and at events, rumors travel faster than the facts. Not everyone who publishes a lot about AI has actually built something that delivers real-world value.

A demo becomes a promise. A promise becomes a trend. A trend becomes a LinkedIn carousel.

And before anyone asks who’s actually done the legwork, it seems like everyone is getting rich.

At the same time, the best AI use cases probably aren’t in those stories. They’re behind the scenes, at companies that use their own data, processes, customer knowledge, and domain expertise to build something others can’t easily copy.

What does that mean for a B2B company with 200 to 2,000 employees? How you sell, what you deliver, your customer data, your insight into who buys what and why, and your perspective on the market—that’s your competitive edge. No tool can provide that. You build it yourself, over the course of years. The companies that will truly derive value from AI in the future are the ones already working on it now.

And for now, there are some players you simply can’t ignore. OpenAI, Anthropic, Google, Microsoft, and Nvidia aren’t just random stalls on the main street. They’re increasingly providing the underlying infrastructure: the models, the computing power, the distribution, the cloud, or the interface that users work with every day.

That’s why now is not the time to fall in love with every single AI tool. The main street of the camp is full of stalls: better shovels, faster sieves, smarter maps, and miracle cures for every conceivable workflow. Some are useful. Some will be gone by tomorrow. Some will be acquired. Some will later turn out to be just another feature in OpenAI, Google, Microsoft, HubSpot, or Salesforce.

We’re still in the early stages—more like a camp than a city. The rules, standards, infrastructure, and consolidation are still largely yet to come.

So what should you do with all this now?

So yes: experiment. Learn. Develop your judgment. See where genuine claims emerge. But don’t commit too early, too deeply, or for too long to individual tools.

Be proactive in learning, but conservative in committing. Especially in a market where today’s tool could be tomorrow’s feature, acquisition target, or footnote.

Oh, and be sure to watch *Deadwood*. Highly recommended!