Hacker News Reader

Comment by ChessviaAI | Hacker News Reader

It feels like we're watching the playbook for AI-native companies emerge in real time.

Duolingo’s approach, explicitly tying headcount to proof-of-automation limits, baking AI usage into performance reviews, and prioritizing AI-first systems over retrofitting old workflows, is a glimpse at how "AI-first" won’t just mean using LLMs as a tool, but rebuilding the entire operational model around them.

That said, it's a double-edged sword. Contract workers were crucial to Duolingo’s early scalability. Shifting to AI removes human bottlenecks, but also human nuance — and teaching language is deeply nuanced. It’ll be fascinating (and maybe a little uncomfortable) to see if mass AI content keeps Duolingo's educational quality high as they chase faster scaling.

AI-first might win on cost and speed. But will it still win on outcomes?

View on HN

> keeps Duolingo's educational quality high as they chase faster scaling

Duolingo is widely regarded as more of a game than a high-quality learning experience. People obvious learn something from it, but it's a running joke almost everywhere on social media that people can be 100s of days into their Duolingo streak and still not learn much.

Getting people off of Duolingo and onto less gamified, more rigorous language learning courses is a common theme in the language learning world.

Any resources you’d recommend?

I haven’t used Duolingo in over a decade, but recently I’ve become interested in learning conversational Spanish.

Language Transfer

Massive plus one for Language Transfer. It's well presented, interesting, and kept me engaged. The whole concept is finding connections to language you already know, and gets you thinking in fuller more complex thoughts and sentences really quickly. The audio lessons are free on various podcast platforms / YouTube etc.

They even explicitly admit to this. In the recent Decoder podcast the CEO said they will always choose engagement and gamification over teaching you the 'best' way.

Which is not a terrible strategy. Most people learning languages are doing it for fun or a new years resolution or whatever. If you're serious about learning a language for real (ie you've moved country) then of course you're gonna go to a more serious platform.

>Duolingo's educational quality high

Was duolingo ever known for high educational quality? To me duolingo's main pitch was a way to gamify language learning. Of course it became a victim of its own success as soon as you could "pay to win".

I don't think so. I see its pitch as "the best kind of exercise is the one you do", maybe preferable to playing a game, but not an efficient way to learn. How useful it is to you will probably depend on how effective the sounds and streaks and home screen notification stuff is for keeping you motivated. Personally, I'm motivated by quick progress and outcomes (streaks don't do anything for me), so Anki is actually stickier, though I must be in the minority.

Because they focus so much on beginning learners for whom nuance isn't important, this change doesn't seem like it'll hurt them.

Being successful at Duolingo was always being like that guy who wins scrabble tournaments in French and Spanish without being able to converse in them. It's just a game and winning at it doesn't necessarily align with being functional in it. Otherwise second language schools would have long been extinct by now.

Far behind are the days when free version of Duolingo was playable. There are so many dark patterns these days to keep users coming back, gatekeeping something or otherwise to just push them to pay for the usage.

I don't think it was ever known for being high quality, it was known for being "accessible" and then they forgot about what their original goals were. They got pretty disappointing IMO when real languages were in need of updates for a long time (I don't know if the Chinese course ever got features like Stories) while they added a bunch of fictional languages.

> AI-first might win on cost and speed. But will it still win on outcomes?

It will be a flop. Either it won't get implemented like the C-levels dreamed in the first place and will remain policy on paper only or it will be rolled back quietly once reality hits.

"AI-first" is the "blockchain" of 2025.

Username checks out, no em-dash needed.

Reply on news.ycombinator.com