If you mean me, I read simple or interesting texts and mark words I recognise as known. No 1,2,3,4,K. It’s simpler and I want to read tonnes, not think about graduations of learning. I used to; I just don’t get much value from it.
Also, the % words read retains a little value for me. If I was to mark each from blue the stat goes to 0% which I find useless. So I just mark it as known or leave it. mY words known is really about having ‘some recognition’ and I amnok with that vagueness.
I imagine people use Lingq is totalltly different ways. The variations could make a reasonable youtube series. Lingq does a decent job. of catering to many different approaches but there are quite a few areas that I imagine people would like improved (without messing it up for other people).
I have never put in an alternative meaning for a word. Again, this lets me get on a read more
Yes, that’s exactly what I’m saying. Blue word - default meaning - adjust level / make known or add additional meaning. In the reader as well as in the vocabulary section. Like Steve did in one of his videos where he scrolls through his vocabulary section and changes the level of words depending on his own assesment.
Thinking back I do this regardless of my level in language.
A “LingQ All New Words” feature would be beyond amazing.
The time saving alone …
What I’m now wondering about the future of AI in language learning is how it could be used not just on the surface—as user-facing tools or experiences—but as a hidden, highly personalized language acquisition diagnostician and content curator. Could AI silently track and adapt to a learner’s precise needs, continually curating material that targets their individual learning edge?
One of the major challenges of the Comprehensible Input method is the sheer volume of content required to achieve fluency. I believe the process could be made far more efficient if the input were meticulously tailored to the learner’s needs.
LingQ does a great job of exposing learners to grammar within authentic contexts. But what if an AI system could go even deeper? Imagine if it didn’t just know the vocabulary I’m familiar with, but could also accurately assess my grasp of grammar—both what I’ve mastered and what I’m missing—by analyzing my input patterns.
Exposure to less frequent grammar points, multi-word expressions, and especially irregular forms tends to take an especially long time with traditional methods. Rather than passively waiting for these to appear in random content, could an AI-enhanced future LingQ identify these gaps and actively generate or curate content that specifically targets them?
For example, if I needed more exposure to the subjunctive mood, could existing content be subtly modified or used to generate derivative versions rich in subjunctive examples—while still maintaining naturalness and comprehensibility?
I’m not suggesting abandoning the core ethos of Comprehensible Input. Rather, I’m asking whether AI could enhance it—elevating it with unprecedented precision and intentionality, reducing the time and input volume needed to reach fluency without compromising natural acquisition.
This language learning site https://morpheem.org/ tries to do some of the stuff you mentioned. LingQ could implement similar features too in the future.
Hey S.m8! I tried Linga and did find korean there. It is just that when I chose Korean, in one go, I had no idea what I was doing as all buttons and menus were in Korean! And not the copy-past-translate type. Everything is non-selectable. So, I never confirmed the email for the tryout.
I would have like to tryout Linga.
I also was wondering about the source of Linga: is it Russian? That would give me cause to pause anything. For the time being.
Exactly! It would basically be like an advanced spaced repetition system—but instead of drilling single words or even isolated phrases, it would generate full texts that naturally include those elements, spaced out in a rational, progressive way over time. It could make the process way more engaging and avoid the boredom that comes with traditional SRS methods
We are adding to our AI team and have some of this in the works. Looking forward to seeing how it works as we start building it out. We are looking for beta testers so let me know if you’re interested. Our first phase is a simple chatbot which we hope to release soon.
Korean is only available on Linga as a dictionary/translation language, but at the moment they don’t have it as a study language. And yes, I understand your reluctance about using a Russian-based app/website. I share it and I wish Lingq would improve the reader dramatically to save me from looking at alternative solutions.
I don’t know what the difference is. What I read about Lingq before joining is that there are new languages (not supported yet) and fully supported languages. Could you explain what the difference is between a dict/translation language and a study language? Does Lingq recognize this difference? Is there docs about it?
Sorry, I wrote Lingq when i meant Linga. My mistake. I have now edited my previous message. Anyway what i meant is LingA doesn’t offer Korean. Only French, Italian, German, Portuguese, English and Russian. You can only set the interface language in Korean and set your dictionaries in Korean. That’s all for now.
Thank you. That clarifies it. Don’t worry about it.
Actually, I am glad to have found someone who is also learning Korean. I was under the impression that not many people do that. I am glad I am wrong. I am currently enjoying the “short end of the hockeystick” and trying to gain vocabulary. Just enjoying the ride.
