This Week at a Glance:

• 💡 Mental Model: Selling shovels in the 1849 Gold Gold Rush

• 📚 Industry Overview: What makes up the AI value chain?

• 📉 Builder’s Log: A personal reflection

1849’s Real Winners Didn’t Mine Gold

A lesson from history about good businesses

In 1848, gold was discovered at Sutter’s Mill in California.

The news spread fast — and by 1849, over 300,000 people flooded west, all hoping to strike it rich. These “forty-niners” came with dreams of fortune, but most left with nothing. Some died broke. Some never left the mines.

But a few people made real, lasting money.

How?

Levi Strauss sold durable workwear.

Samuel Brannan bought every pickaxe in San Francisco and jacked up prices.

Merchants, hoteliers, laundromats, and transport companies quietly built empires while others chased dreams in the dirt.

This is the shovel seller’s edge — and it’s one of the most powerful mental models in business.

Don’t chase the boom. Own what the boom needs to function.

Shovel sellers don’t need to guess who wins. They profit from the attempt.

It’s a pattern that repeats every time:

• In a gold rush, sell mining gear.

• In a real estate boom, sell title insurance and software.

• In an ecommerce wave, sell payment rails and logistics.

• In AI? Well…we’ll get to that.

This isn’t just clever — it’s a more resilient way to allocate capital.

• Shovel sellers have broad exposure, not binary outcomes.

• They win if the category grows — regardless of who takes the spotlight.

• They often have high margins, recurring revenue, and strong moats.

For investors, this mindset is gold.

Find the trend. Then ask: “What does every player in this trend depend on?”

That’s where the real money is.

AI is Booming

So what are the easiest ways to profit?

The gold rush is here again. This time, it’s artificial intelligence.

Billions are pouring into LLMs, AI copilots, content tools, and autonomous agents. Founders are racing. VCs are betting. Everyone wants to strike gold.

But most of the attention is focused on the top of the stack — the apps.

Just like in 1849, the real opportunity might not be in who builds the “best” AI product

It’s in who sells the shovels to all of them.

Let’s walk the AI value chain — from the foundation up:

🧱 1. Chips (Physical Compute)

The shovel factory.

• Every model needs GPUs — and almost all roads lead to Nvidia.

• Other players: AMD, TSMC, ASML (manufacturing and lithography)

• These companies profit before a single AI app reaches a user

☁️ 2. Hyperscalers (Cloud Infrastructure)

The landlords.

• Models don’t run on laptops — they need scalable cloud hosting.

• Think: AWS (Amazon), Azure (Microsoft), GCP (Google)

• Even OpenAI runs on Azure — and pays heavily for it

🧠 3. Data Infrastructure & AI Platforms

The control panels.

• Enterprises don’t just need models — they need clean data, pipelines, and governance

• Key players: Snowflake, Databricks, MongoDB, Hugging Face

• These firms win when companies do something with AI — at scale

📱 4. Applications (What People See)

The miners.

• These are the UIs, apps, and experiences on top: ChatGPT, Perplexity, Jasper, etc.

• They get the headlines — but face brutal competition and low switching costs

• Many are just wrappers on APIs powered by layers 1–3

💡 Bottom line?

Most people try to pick the winning app.

Smart money bets on the stack.

It’s easier to bet on the infrastructure layer that gets paid every time someone tries to build something in AI — than it is to guess which app will survive next year.

These are the shovel sellers of the AI age.

From User to Builder

A short reflection on my entrepreneurial journey

The concept of selling shovels is easy to understand, but harder to break free of.

I use shovel sellers every day:

• Beehiiv for my newsletter

• TikTok, YouTube, Facebook, LinkedIn to distribute content

• Fidelity and public exchanges to invest in businesses

I’m the miner.
I’m the user.
I’m the customer.

And that’s fine — for now.

But it got me thinking…

What would it look like to become a shovel seller myself?

How could I build infrastructure instead of just using it?
How could I sell tools, not just content?
How could I create something that scales others — and gets stronger as they grow?

It’s a different kind of leverage and I’m not there yet. But this question is shaping my future plans.

For now, I’ll keep digging.

But I’m also sketching out my own shovels.