Live data from Hacker News

Duolingo CEO tries to walk back AI-first comments, fails

htxt.co.za

331–340 of 348 posts

Re: Duolingo CEO tries to walk back AI-first comments, fails

#331

Earlier quoted context omitted.

I disagree with this in principle. Gamification is something we should be very wary of because it is inherently bad. It reduces what you care about in an activity to points and a progress bar. Instead of sticking with language learning because you have some intrinsic reason to want to learn it (or even a external one such as wanting a new job) you're substituting that with whatever Duolingo puts for their gamificatio…

I have to say I think this is a to each their own type of thing. If the goal is to learn a language there’s no extra credit for your motivations or drive. There’s no uber mensch superiority between a person who leveraged gamification to practice or someone who steeled themselves with a few pages of Nietzsche before settling into a determined five hour rote study session. I personally find the gamification of Duolingo…

I think it's more that the process of actually learning a language is much more time consuming than people expect. There's a sort of idea if you haven't done it before that you learn all the grammar rules, learn translations of each word, and then you're good.

But once you get into it you realise that doing this literally is both a massive task and also Sisyphean. Learning it like an algorithm you run through in your mind is way too slow. You have to just listen and interact with it enough that some other strange alien part of your mind can remember and understand it implicitly. It's a weird process to experience, when things "sound right" without knowing why etc.

Duolingo with it's single sentences, and especially how it tries to have 1-1 translations for everything is good, but not sufficient. It's best used as just a part of study, something to get vocab and a sense of basic grammar rules.

Why is it bad? Because their approach hits a plateau. The gamification at that point goes from good to bad. It's why people switch languages and learn like four at a time in Duolingo (bad for the alien brain bits that are trying to develop the one language) rather than stick with the one (because beyond a point it's more of the same level rather than moving up at a good pace)

It also kinda takes you away from going outside Duolingo and seeking out other things because if the gamification works you're doing that plus Duolingo rather than doing what's actually appropriate for your level

Re: Duolingo CEO tries to walk back AI-first comments, fails

#332
post #252

Earlier quoted context omitted.

I don't agree with the person above you in as much as the way they are doing it is very individualist "vote with your wallet", and yes, you're right that it's very ineffective. The more effective way is to form a group, call it a "club" or whatever, that does it. The group can then advertise to other people and get more people to join the club. Eventually, it becomes large enough to gain political power. This is call…

Influencing is far more powerful than unionizing. And activism is an ineffective form of influencing.

In the UK, McDonalds workers unionized, and now they get paid 15£/hr.

The problem with Unionizing in the US, is there's very little cultural problem with crossing picket lines.

Re: Duolingo CEO tries to walk back AI-first comments, fails

#333

Earlier quoted context omitted.

Because you "walk into" things and get hurt. Like walls, or trees. The notable exception is buildings, or less tangible things like a spotlight. Trains, ships, and busses are a weird kind of middle ground between on/in. Thinking about it, I believe the principal difference is if there is something significant (more significant than the roof of a train) above your head? It seems to be very subjective and though, any t…

In drugs, on only my socks.

Underrated comment.

Re: Duolingo CEO tries to walk back AI-first comments, fails

#334
post #293
post #173

Earlier quoted context omitted.

The quality of advice on Reddit seems to be variable, unfortunately. Apparently https://www.reddit.com/r/LearnJapanese is full of people who've spent years doing Genki ( https://genki3.japantimes.co.jp/en/ , probably an even bigger sinkhole for money, time and enthusiasm than DuoLingo) with little to show for it who are determined to drag everyone else into the same crab-bucket, for instance. OTOH https://www.reddit.…

What do you recommend?

For Japanese specifically? Try starting with https://refold.la/how-to-learn-japanese/ . (There are various people promoting similar methods, but Refold are probably by far the most pleasant and professional option.) And in general https://refold.la/get-started/ and the Refold discords are usually a good starting point. For languages which have Muzzy https://en.wikipedia.org/wiki/Muzzy_in_Gondoland translations I'd honestly just start by watching Muzzy in Gondoland maybe thirty or forty times, then moving on to Muzzy Comes Back, before worrying about any other resources or methods. Spanish is unusually well provided for, with Dreaming Spanish https://www.dreamingspanish.com/ and also Destinos https://www.learner.org/series/destinos-an-introduction-to-s... .

Re: Duolingo CEO tries to walk back AI-first comments, fails

#335
post #149

Earlier quoted context omitted.

So I think people who grew up in Europe surrounded by many languages have a huge language learning advantage over mono-lingual Americans. Also exercises like "get a pen and paper and learn vocab by repeatedly writing it down" this is exactly what language applications have you do, just minus the pen and paper. (I'm not a young person so I have deep familiarity with pens and paper)

In my own personal experience, the pen and paper is actually a very important tool of memory to which tapping on a phone doesn't compare.

Opposite for me, I struggled with courses that required thorough notes until college where they didn't care if I brought in a laptop. Once I could take digital notes I excelled thanks to the speed and organization of it. I don't think the physical instrument for note-taking matters so much as the act of note-taking.

Re: Duolingo CEO tries to walk back AI-first comments, fails

#336
post #8

Their comments cemented to me that they have no long term value. If the ceo of Duolingo thinks AI will teach me a language then I’ll use a low cost LLM to get there without Duolingo.

I've noticed that a strong thread of the hacker community, including HN and the guys who wrote GNU and linux, are extremely cheap . Like, you'd rather write a product clone yourself than pay $15/month for the official product. Why is this? What's with the inherent stinginess?

You say that like saving money is somehow a bad thing.

Re: Duolingo CEO tries to walk back AI-first comments, fails

#337

Earlier quoted context omitted.

Because you "walk into" things and get hurt. Like walls, or trees. The notable exception is buildings, or less tangible things like a spotlight. Trains, ships, and busses are a weird kind of middle ground between on/in. Thinking about it, I believe the principal difference is if there is something significant (more significant than the roof of a train) above your head? It seems to be very subjective and though, any t…

In drugs, on only my socks.

But the feet are in the socks and the drugs are inside you!

From the socks perspective they are on you!

Re: Duolingo CEO tries to walk back AI-first comments, fails

#339

The app has gone down hill but i still use it because i cant find any decent Irish programs elsewhere. I dont mind the gamification but can see why its annoying. Once I fond a suitable replacement, i will cancel my membership. I will say this, ive learned quite a lot of the language simply through the app. I am highly motivated though.

Watch these over and over for a month: https://youtu.be/496H4TdJIiw

Re: Duolingo CEO tries to walk back AI-first comments, fails

#340
post #56

Earlier quoted context omitted.

I do think AI could be a better language tutor than an average language teacher, but I don't think Duolingo's approach is very effective. The ideal AI-powered tutor would work more closely to a private language tutor. It would speak with you and gradually integrating language concepts into the conversation. When you make mistakes, it could correct you on the spot and keep track of where your strengths and weaknesses…

> I do think AI could be a better language tutor than an average language teacher, but I don't think Duolingo's approach is very effective. No offense, I can tell you're neither a linguist nor a language teacher. This is one are where the human input is invaluable and irreplaceable. Because language (the complex kind) is inherently human. It quite literally is.

I'm neither, but I tried learning Spanish from a "language teacher" and it didn't work. I'm not alone. I've spoke to dozens who had the same experience.

Years later I learned from people with the same method I described.

The AI models (large LANGUAGE models) of today seem very close to replicating that experience.

Try telling ChatGPT with voice mode "let us talk about X in [pick language]... Correct me when I'm wrong and explain why".

I'm curious where you think it's failing.

What do you feel is missing to replicate a good language teacher?

Post reply on HN