Interesting. I've been working on the same problem (you need stuff at a kid's level, but not aimed at kids, if you are interested in the content then it will be easier to learn): https://tailoredreaders.com/samples
I think I like your solution better than mine, which is basically generating children's books from user prompts.
When you say you generate stories from the films, are you summarising / simplifying the narrative of scenes in the film (to down level the language), or creating something new, just with the vocabulary the film uses?
It's a similar methodology, my problem with things like this is that they do give you comprehensible input, but it's very boring. I don't engage at all. The idea of Katarineko is to create comprehensible input from content that interests you. Say you want to watch Spirited Away but it's a bit too advanced for you at the moment, we create stories at your level for Spirited Away to teach you the core words you need to achieve your target comprehension level.
Depending on my mental state, status with the person and many other factors each of them may trigger both many different internal thoughts, looks, body expressions and even outcomes.
I find that my orchestrating agent still needs to be fast to properly coordinate a fleet of slower models.
During post-training of opus 5, the last few days, opus was a real wreck. I had to swap in gpt 5.6 sol for my orchestrator and enable fast mode (1.5x speed) in order for it to keep up with work and communications from a handful of mostly 5.6 sol agents.
Also because interacting with a slow orchestrator is no fun, even when plenty of work is getting done in parallel in the background.
I've been building a board game community on ATProto. The idea revolves in trying to build the feeling of local clubs but online, instead of "PLAY NOW" being on the landing page (although you can play now if you want) it's organized around clubs. I also wanted to move away from one game oriented community. I play Go and Chess and I would love clubs to be for mixed games. You can create leagues, or tournaments, etc.
As an extension of that I've been building a small language to build games (with a nice time traveling debugger and automatic replays) inspired by Elm. Go and Chess are both built on top of this and you could fork them and modify the rules. Want to code a new Chess variant? Go ahead! Want to code a whole new game you want to try? Go ahead!
What I'm very excited with ATProto is that because all games are broadcast on it and each shares a core system of turn-based replays (and review system with branching) software can be built on top to do AI analysis, or a replay view. The idea that my application can be extended without me adding an API is very exciting and very much in the direction I want to take the system (community is the core tenet).
So far they are mostly trusted authority rather than peer systems. But, since each player owns their own PDS state, one could perhaps just hash the previous state they are working off of. That can at least detect if a game is consistent/agreed to by both sides, allow one player to detect the other's tampering.
This would imply the game state is stored in a centralized app view, at which point the general atproto ethos that you can move your data if an app becomes bad is up to the app.
I believe the viable solution in a rich permission system is one where the user cannot directly modify a record (as defined by a permission schema for a specific app/type) and can only call methods.
This can easily just be gentle ribbing where the juggler and "critic" are familiar/friendly enough that the juggler understands the dude respects his juggling skills and knows the critic is just making friendly banter when he sees him making more frequent drops.
And yet my American ex-GF that went to one of these top universities with very good grades when she came to Barcelona to do a masters (in a lowly Spanish university) she was so far behind in knowledge compared to her peers she had to do a lot of work just to catch up.
I'm pretty sure most of these are just politics being played.
We already have instant SEPA payments and most countries already have a version of this running, I think the idea is that this unifies it across europe. I can send money to any friend in 20 seconds with Bizum in Spain and a lot of merchants also accept Bizum.
I've been learning Japanese as well and the content at my level is boring while the content I'm interested in is a wall.
reply