Duolingo’s stock is down 72% in a year, and it might be one of the most important lessons from AI right now!
The company that did everything right on paper.
Duolingo embraced AI early:
- AI-generated lessons
- AI-powered personalisation
- Automation replacing parts of human content creation
By the usual narrative, this should be the future.
So why the brutal drop in their valuation?
Because the AI market isn’t rewarding adoption anymore. It’s rewarding defensibility.
And AI lowers barriers to entry. When everyone has access to the same models, the question shifts from:
“Are you using AI?”
to:
“What do you uniquely own that AI can’t commoditise?”
That’s where things get uncomfortable.
What actually creates defensibility now?
- Proprietary data (that others can’t replicate)
- Distribution that compounds
- Brand tied to outcomes, not features
- Community with accountability
- Execution speed inside a real operating model
Here’s the trap we are seeing everywhere now – AI enthusiasm being mistaken for AI advantage. They are not the same.
- AI is a capability.
- Not a moat.
- Not a strategy.
- Not a business model.
Enduring value comes from how AI is integrated:
- into business workflows
- into human accountability
- Into structures and measurable outcomes
- into culture and execution
If a business is built on surface-level differentiation, AI strips it bare. If it’s built on real operations, evidence, and execution, AI quietly compounds the advantage.
The lesson isn’t that AI failed the business. It’s that AI amplifies foundation either strong or fragile.
That distinction is about to matter a lot more than who adopted AI first.
#AIReality #RealMoats #BuiltToLast #OperationalExcellence #BeyondTheHype #SystemsOverTools #educli