pull down to refresh
27 sats \ 6 replies \ @kr OP 23h \ parent \ on: The Duolingo Handbook tech
The size of the education market being dramatically underestimated because historically it was paid for via our taxes.
Americans spend $10k+/student/year for pretty bad education outcomes.
What do Americans spend on Apple products? $1k/year?
How much revenue do Americans generate for FB/Google? Probably a combined $600-$800/year
Huh. It's big, but also well-served at every price point, including zero; and especially well-served at $20/mo, racing to zero.
I will stay tuned.
reply
reply
Nothing in their current offering seems like a moat to me, not in the sense that the word is usually used. Scaling a bunch of human contributors, and enjoying some modest network effects from that, could have been a moat. But then LLMs got invented.
If they want to make "the best way to learn stuff" truly general (e.g., calculus; coding; dating), and they had a secret sauce that differentiated them from the other players in those spaces, then I'd be very mildly interested -- still seems like learning chemistry or whatever could improve by an order of magnitude, and I'd pay for that. That's a tall order, though.
reply
The 1T dollar number you put up kept popping up in my head during breakfast.
I don't believe Duolingo will ever reach that. They have too much of an altruistic mindset to belong in the same group of companies that currently have a 1T marketcap. The language certification they provide as alternative to established but expensive other certificates is an example of this altruistic modus operandi.
At the same time, what several current 1T companies represent (walled gardens, bastions of censorship, etc) is what we've come to despise in recent years, so I don't think this should be a metric in itself.
Things can change of course.
reply
I also don't think they could ever get to $1T without going back to community sourcing. They'd have to branch out into all sorts of other fields with different teaching and assessment needs, would be very hard and expensive to do in-house.
reply
Exactly. Curriculum design is really hard. And if you branch into other areas, now you have real competitors. And if the secret is using LLMs, well, lots of people have way more horsepower to do that than Duolingo. The fact that they haven't suggests it's a boutique area without much money to be made there.
reply