I did 98,000 Anki reviews. Anki is already dead

Anki helps the best when you are at the beginning of the language journey. With a vocabulary of 0 words you won't be able to read anything at all, and there's nothing to talk to an LLM about yet.

Those 'useless' static cards are extremely efficient for learning the 1st 2000-3000 words, which is key to start reading. After about 4000 there's little sense in using SRS anymore, and then I'd rather spend more time with an actual book, but getting there with anki felt like using a cheat code compared to how I learnt my first foreign language. It's not exciting, it's pure toil, but it does work.

And when it comes to the next stage, I can't imagine how random llm-generated texts are better than, say, graded readers or real books. Most people would likely find it more interesting to spend an hour or two a day following an exciting story and characters they care about, and it's (based on a sample of one) way easier to memorise all of those new words when there's an emotional connection for each one (just how we form associations between words and experiences while growing up).

As for the app itself- I have tried it with my native language, and at the advanced level it produced a sterile and slightly unnatural text with a complexity of a typical fiction. If someone could read this, then I don't see why they would bother. At the beginner level the app generated a couple of news stories which, though simple grammatically, had a vocab that I would never have recommended to a novice. Local news of a "a firefighter saved a kitten stuck on a tree" variety are much more useful for that kind of learning, and you get this from any free newspaper.

LLMs are extremely useful for learning foreign languages, but I feel like this isn't the way to go

In my limited experience with static cards you really do just end up learning the "appearance" of a card, which is why I've been more successful with dynamic cards, like say a card showing a concordance randomly selected from a large corpus every review, showing the word I want to learn in various contexts.

I agree that sentences are generally superior to vocab. Vocab cards are extremely problematic once you go beyond beginner level, because words can (and often do) have multiple translations in both directions. This could be because there are multiple words meaning the same thing, or because a word has multiple meanings.

For example if the English prompt is "watermelon" - are you supposed to recall the Italian word cocomero, anguria, or melone d'aqua (all of which mean watermelon)? If the English prompt is "bank" - is that a place you deposit money, a river bank, to bank (turn) a plane, or to bank (count) on something happening? You end up having to build in messy hacks like giving clues in the prompt as to which translation is intended (which means you memorise the clue instead of the word) or having cards for bank(1), bank(2), bank(3), and bank(4) which becomes very tedious for recall. Sentences mitigate these problems somewhat.

I now only use vocab cards for object nouns where there's only one important translation, and mainly because I can put pictures on these cards so that I'm learning from e.g. the concept of an orange instead of the English word for orange (which saves you the step of mentally translating when you aren't yet fluent with the word).

Anki is dead because, after LLMs were introduced, people lost the incentive to learn at all.

Frankly, who wants to learn something the old way when you can ask an LLM to do it for you?

All this reminds me of my home country when the Ponzi schemes were at their highest point (2008). There were small and medium cities, were a sizeable percent of the population actually stopped working at all! Why? Because the ponzi schemes were promising up to 300% returns within 6 months! And the Ponzi schemes were actually delivering (so much that the banks actually pressured the government to intervene). So people put all the money in the ponzis and simply wait a bit to cash in part of their returns. No one opened their small grocery stores, nor small pubs, nor did anyone want to do chores or manual labor. Until everything collapsed.

I agree that they could be complementary, but I think there's a not-yet-made tool that goes even beyond this, where you are interacting with an LLM that has an Anki-like backend of some kind, keeping track not only the number of mistakes but of what kind of mistakes you made and when, so that it can later bringing up the card in a more natural way.