The Cargo Cult of Silicon Valley

I was in a meeting last week where a VP said, with complete sincerity: "We need AI in our product. I don't know what it does yet, but every competitor has it."

And I had the weirdest déjà vu.

Because I'd heard this exact reasoning before—not in a boardroom, but in an anthropology class twenty years ago. The professor was explaining cargo cults in the South Pacific after World War II.

And holy shit, the parallels are eerie.

What the Hell is a Cargo Cult?

During WWII, the US military built bases on remote Pacific islands. They brought planes, radios, jeeps, medicine, food—cargo. The indigenous islanders watched soldiers perform strange rituals:

• Wearing headphones and talking to invisible people
• Waving colored paddles at planes
• Building towers with blinking lights
• Creating landing strips in the jungle

Then the soldiers would do these rituals, and planes would land carrying miraculous goods.

Makes sense, right? Cause and effect.

After the war ended, the soldiers left. The planes stopped coming. The cargo stopped arriving.

So what did the islanders do?

They recreated the rituals.

They built fake runways out of bamboo. Carved headphones from coconuts. Waved palm fronds at the sky. Built wooden control towers with torches for "lights." Some even marched in formation wearing hand-carved "uniforms."

They perfectly mimicked the form while completely missing the substance.

And they waited. And waited. And the planes never came.

This isn't a story about "primitive people" being naive. This is a story about rational humans observing a system they don't understand and doing their absolute best to replicate the visible parts.

Sound familiar?

The AI Cargo Cult of 2025

Walk through any tech company right now and you'll see it:

Company A sees Company B deploy "AI chatbots" and get funding.
So Company A builds a chatbot. Doesn't matter if their customers need it. Doesn't matter if it solves a problem. They saw the ritual. They're copying the form.

Startup C sees OpenAI's ChatGPT succeed.
So they wrap GPT-4 in a thin UI, add some prompt engineering, and call it "revolutionary AI for [industry]." They've built the bamboo runway.

Enterprise D sees competitors mention "machine learning" in earnings calls.
So they hire a "Head of AI" and start every meeting with "How can we add AI to this?" They're waving the palm fronds.

Nobody knows why the planes are landing. But everyone's building runways.

The Thing About Real Runways

Here's what the cargo cultists couldn't see:

The planes didn't come because of the runways. The runways existed because planes were already coming. The military built bases where they needed supply lines. The rituals were documentation of a working system, not the cause of it.

The headphones connected to real radios. The control towers had actual radar. The landing strips were engineered for specific aircraft weights and speeds. The cargo existed because there was demand, supply chains, and logistics infrastructure spanning continents.

You can't summon a supply chain by cosplaying one.

What Real AI Success Looks Like (Hint: It's Boring)

Let me tell you about two companies I've worked with recently.

Guess which one talks about AI in their board meetings?

Company 1 shows up at conferences with slides about "AI transformation." Company 2 quietly saves $2M/year and doesn't mention AI at all—they just say "we improved our support system."

Company 1 built a runway. Company 2 fixed a real problem.

The Rituals vs. The Reality

Here's the cargo cult checklist for 2025:

See the pattern?

Cargo cult: Start with the tool, find a problem.
Real engineering: Start with the problem, find the right tool.

Why Smart People Build Bamboo Runways

This isn't about stupid people doing stupid things. The cargo cultists were smart. They observed patterns. They replicated behavior. They were rational.

They just didn't have access to the underlying system.

Same thing is happening in tech right now:

• VCs are funding "AI companies" → Companies add AI to get funding
• News talks about AI disruption → Execs panic and demand "AI strategy"
• Competitors announce AI features → Product teams scramble to ship AI features

Everyone sees the surface-level rituals. Few understand the actual mechanics.

It's not malicious. It's not even irrational. It's pattern matching without system knowledge.

The Questions Nobody Wants to Answer

Want to know if you're in a cargo cult? Ask these questions:

When the Planes Actually Land

Here's the thing: Sometimes AI is the right answer.

I refactored a 2,000-line Express.js API in 8 hours with Claude Code. Would've taken 2-3 weeks manually. That's not cargo cult behavior—that's using the right tool for the job.

A client used AI to analyze 10 years of customer support tickets, found patterns nobody saw, redesigned their onboarding flow. Reduced churn by 15%. That's not a ritual—that's solving a real problem.

Another company uses AI to generate first drafts of technical documentation. Engineers review and edit, but it cuts doc time from 4 hours to 45 minutes per feature. Not magical. Not revolutionary. Just the right tool for a specific job.

The Uncomfortable Truth

Most "AI transformations" are cargo cults because most companies don't actually have AI-solvable problems—or if they do, they don't know how to identify them.

It's easier to:

• • Copy what successful companies are doing
• • Add "AI" to get funding/press
• • Follow the herd
• • Build the visible parts of a system

Than to:

• • Deeply understand your bottlenecks
• • Admit which problems you actually have
• • Do the hard work of integration
• • Measure results honestly

Building bamboo runways is easier than building actual supply chains.

How to Avoid the Cult

If you're leading a team/company and want to avoid cargo cult AI:

1. Ban the phrase "AI strategy" until you can list 5 specific problems you're trying to solve
2. Start with pain, not tech: What's slow? What's expensive? What's error-prone?
3. Measure before and after: If you can't measure improvement, you're performing rituals
4. Kill projects that don't deliver: Cargo cults never abandon the ritual even when planes don't come
5. Hire skeptics: If everyone's excited about AI, you're in an echo chamber

The Lesson from the Islands

Eventually, some islanders figured it out.

They realized the runways didn't summon planes—planes came because there was trade, because there were resources, because there was demand. The rituals were symptoms, not causes.

The same realization is coming to AI.

Companies will figure out that "having AI" doesn't create value. Solving problems creates value. AI is just one tool among many.

The cargo cults will fade. The planes will land for companies that built real infrastructure, solved real problems, and created real value.

And everyone else will be left holding bamboo headphones, wondering why their product-market fit never materialized.

The Bottom Line

If you're building AI features because your competitors have AI features, you're in a cargo cult.

If you're building AI features because you identified a specific, measurable problem that AI solves better than alternatives—congratulations, you're doing engineering.

The planes don't care about your runway. They land where there's cargo to deliver.

Stop building runways. Start creating value.

And if you can't tell the difference? You might want to take off those coconut headphones and look around.