Hey folks,

Before we dive into how a green owl made crazy money, here’s some good stuff from the world of AI.

Curated Chaos

1. Want a list of AI tools no one is talking about? - Click here

2. Want to watch a quick video on AI updates of the week? - Here you go!

3. I recently watched a TEDtalk. Sam Altman discussing the scariest thing about AI. Got me hooked. - Watch it here

4. Prompt of the week? The Grandma Exploit. Yep, that’s the name. It’s as interesting as it sounds. - Read more here

Talking of interesting stuff, let’s look at some news that matters.

AI Intel of the week

This week, something caught my eye that I had to dig into:

and made a ton of money doing it.

You know that green owl that passive-aggressively reminds you about your Spanish lessons?

Turns out, there's a lot more going on behind the scenes than most people realize.

Some numbers so you care:

• Revenue jumped from $531M to $748M in one year (+40.8%)

• 34.1 million people use it EVERY DAY (that's more than the population of Canada)

• They're projecting to hit almost $1B in 2025

• 82% of their revenue comes from subscriptions.

But here's what really got my attention: While everyone else is slapping "AI-powered" on their features and calling it a day, Duolingo figured out how to get people to pay $30/month for AI tutoring.

Let me walk you through exactly how they did it.

The backstory (you might relate):

Luis von Ahn wrote his graduate thesis in 2005 about AI helping humanity. 2005! While we were all using flip phones, this guy was already thinking about AI-powered education.

Fast forward to 2020: ChatGPT drops, and Duolingo went absolutely bananas with their AI strategy.

The 3-step playbook that changed everything

Step 1: The Content Factory Revolution (2021-2023)

What everyone thinks they did: "Oh, they just used AI to make courses faster."

What they actually did: Rebuilt their entire content creation machine from scratch.

Here's the genius part:

The result? They created 148 new language courses in under a year.
For context, their first 100 courses took 12 years.

Key insight: They didn't replace humans immediately. AI became the first-draft creator, humans became the editors. This let them scale 10x while keeping quality high.

Step 2: The Premium Experience Masterstroke (2023-2024)

This is where the money magic happened.

Instead of making AI features free (like literally everyone expected), they launched Duolingo Max at $30/month.

The AI features that justified the premium:

"Explain My Answer" → When you mess up, GPT-4 breaks down exactly why in normal human language

"Roleplay" → Practice conversations with AI so you can embarrass yourself safely before talking to real humans (josh talks model)

"Video Call with Lily" → This one broke my brain. You can video chat with their animated mascot. It's like FaceTime with a patient, encouraging AI tutor.

Luis nailed it: "We wanted tutors as effective as humans and as fun as Candy Crush."

The psychology here is brilliant. Language learning is scary because you're afraid to make mistakes. AI removes that fear, you can embarrass yourself in front of a machine without feeling judged.

Step 3: The "AI-First" Controversy (2024-Present)

This is where things got controversial.

Von Ahn announced they're now "AI-first," which includes:

The math is harsh but simple: When AI can generate lesson drafts 100x faster than humans, you need fewer humans.

The competitive advantage: 130 million users generate billions of learning data points. Every mistake, every pause, every success feeds their AI models. That's years of behavioral data that competitors can't replicate.

Why this matters for you

Three takeaways I'm stealing for my own work:

1. Don't just add AI features—rebuild the entire workflow Duolingo didn't slap ChatGPT onto their existing process. They reimagined content creation from the ground up.

2. Premium AI experiences can justify premium pricing. While everyone gives away AI features for free, Duolingo proved people will pay $30/month for AI that solves real problems.

3. Data moats are everything in AI Their 130 million users create an insurmountable advantage. Every interaction makes their AI smarter while competitors start from zero.

The uncomfortable truth

With 56% market share in online language learning, Duolingo is becoming a monopoly.

The AI advantage compounds: more users → more data → better AI → better experience → more users.

It's the ultimate flywheel, and they figured it out while everyone else was still debating whether AI is a fad.

What do you think?

Are we seeing the blueprint for how AI companies will dominate their categories?

Want more breakdowns of how AI has changed the path for companies?

Let me know here.

Closing thoughts

What happens when AI gets so good that we prefer it to humans?

Duolingo's AI tutor is more patient than most teachers.

Cisco's AI customer service will be more helpful than most call centers.

And if Anthropic can now legally train on all human knowledge without compensation...the question isn't whether AI will change everything.

The question is: Are you building the future, or is the future being built around you?

Hit reply and tell me the most unique thing you have done using AI.

The best response gets a shoutout in the next edition.

Until next time,
Vaibhav 🤝