- "all" might be a bit of an exaggeration, but the philosophy is to learn to recognize roughly 2000 kanji before starting the actual language learning. Volume 2 and 3 are supposed to complement more normal language learning.
The theory is based on the authors experience seeing Chinese and Korean students learn much, much faster than their western peers in Japanese language classes, coupled with an argument for "If you can read 50% of characters, you still can't read"
I'm surprised you've never come across this, as it is in the foreword.
> There will be somebody, somewhere who took this path and is happy to tell you about their success using it.
I met this somebody in Japan. If I remember correctly, he spend a summer "doing" RTK, then took 1 semester Japanese at his home university, went on exchange to Japan for two semesters, and after finishing his first semester abroad he passed JLPT 2 (not N2 - this was before they added the N)
Good for him. He was a strong student, but I wouldn't recommend it.
- There’s another school of teaching, where kana and kanji are banned for the first 2-3 semesters because they are a distraction to learn and internalize words and grammar.
I’ve met a few students of this textbook system when I was on exchange and my impression was that they were very skilled at Japanese for the amount of time they’ve been a student and what they told about their seniors was they pick up kanji fast, since they already know the words.
The big problem of course is that it is completely incompatible with other schools. Where do you place them when they go on exchange? With the n3 or n5 students?
Anyway, I always thought it was interesting that the exact antithesis of RTK* exists and works.
*RTK or “remembering the kanji” is a system that teaches all kanji before student learn their first word. It’s quite popular online as it lends itself very well to solo studying.
- No, their millions of dollars dont come mostly from people wanting to support their browser.
It comes from search ads on google.com
- Isn’t X.T still valid? I believe it’s been phased out in pandas but still around in numpy.
- Sorta depends on where one draws the line, but the line you are drawing, suggest there should be no government, so "system not working as intended"
- I’d like to have deck-wide variables/lookup tables and links.
The decks for studying Japanese that I’d like would have RTK/wanikani style elements used for mnemonics and I’d like them shown in the answer along with a full description and cross references.
Right now I’d have to build a templating system to prebuilt my deck and import it and it’s just a lot of work on top of the work of building the content, but mostly it makes it difficult to edit/update cards while studying.
- > First of all, the promise of easy portability breaks as soon as the script has dependencies.
And bash has a good dependency story? At least with python you can bundle your script with a requirements.txt file and it is doable for the target machine to get up and running.
- How does this compare to similar schemes in other countries? I believe Portugal is a popular destination for acquiring EU Passport. The UAE also has a golden visa, but it is not a path to citizenship or a passport.
- Lots of people complaining that this has warped textures and whatnot - but come on! This is amazing!
- Is it correctly understood that this is Anki for a subset of leetcode problems with study notes?
I bit more info on what NeetCode is, why I should focus on those 150 problems and how the drilling actually work would be helpful. Do I get asked to do the same problems on repeat? Is it the same problems reformulated over and over? Is there actualy any spaced repetition, or am I projecting?
- This is why I’m surprised by headlines like this”nobody wants to by Microsoft’s AI” like, every corporate M365 user must be either considering it or already started the purchasing process.
- You have it backwards. Building a flexible system and constantly changing pricing database without regard to how to physically update prices in the store is the silly thing.
And when the mismatch tends to be in the stores favor, then maybe it isn’t silly but malicious.
- I started reading this because of your comment. Maybe someday I’ll recommend it in a HN thread and some unsuspecting HN reader will come to read it too!
- You can send the phone number to yourself in chat and then click it to open it for chat.
- Oh… I’ve never done these “in time” and wondered why the next day problem felt so much easier.
I always put it down to overthinking and never arriving at a solution but maybe it was actually a much tougher problem!
- I don't feel like this entry is harsh on authors father. I think it is nuanced and forgiving. It seems most commenters disagree with me on that.
Regardless, it is a well written article with an emotionally strong impact. Thank you for sharing.
- I'd say one of the main differences is that a Markov chain trained over N-grams works on discreet n-grams. Therefore the markov chain cannot tell the difference between two contexts never seen in training. They will both be the "unknown"-token.
An LLM will see a bunch of smaller tokens in a novel order and interpret that.
- The first chatgpt models were kept away from public and academics because they were too dangerous to handle.
Yes it is a thing.
- A video of the start of his latest lecture where he talks about this is trending on reddit (and presumably TikTok et al.)
But still impressive.