I built https://rebrain.gg. It's a website which is intended to help you learn new things.
I built it for two reasons:
1. To play around with different ways of interacting with a LLM. Instead of a standard chat conversation, the LLM returns question forms the user can directly interact with (and use to continue the conversation with the LLM).
2. Because I thought it would be cool to have a site dedicated to interactive educational content instead of purely consuming content (which I do too much).
An example of a (useful-for-me) interactive conversation is: https://rebrain.gg/conversations/6. In it I'm learning how to use the `find` bash command. (Who ever knew to exclude a directory from a look-up you need to do `find . -path <path> -exclude -o <what you want to look for>`, where `-o` stands for "otherwise"!)
Still very early on, so interested in and open to any feedback.
Thanks!
For feedback:
* Please change the colours. I can't read white text on light purple. By all means keep purple, just increase the contrast.
* Maybe keep it in a narrower column, again for readability especially since we're constantly switching sides left to right as we read the AI text and the 'You submitted' text. I also feel it's quite duplicated, I can visually see which option was chosen yet it's restated in 'You submitted'. Only show what's necessary.
* In the replay you shared I can still change active radio buttons. Make it static.
To more focused feedback:
* I really like the idea. Instead of an AI telling you the answer, you learn -- this is fantastic.
* Multiple choice questions are great, but eventually please move to freeform. An AI can evaluate that. (For example, after learning a bit via examples, it might prompt, 'How do you find all PDFs in your home folder?' and give a text entry field. In the past, you'd have to type exactly the string the app expected; with an AI, it can evaluate flexibly. So take advantage of this. It's an immense opportunity vs earlier learning platforms.) Also, people learn better at some point when you're not given answers (aka multiple choice) but need to write.
I think you're on to something. AI as thinking replacement is a worry. AI as a guided teacher is great. I look forward to what you do with it.
Re the more focused feedback, I totally agree re the questioning styles. In the prompt I ask for it to not do so many multiple choice questions, but I think it is addicted/the conversation history skews the context.
I'm going to introduce a settings panel (easily accessible during the conversation), which will let you move to "chat mode" (to discuss instead of be asked questions), and also to configure the types of questions you're asked and the ratio (if I can get the llm to oblige). I'm also going to see if I can come up with some different question formats beyond multiple choice, free-form and multi select (which the llm doesn't use too much).
One potential direction, simple cards that are True/False.
Directed learning towards some goal is always going to be useful. These replacements for doom scrolling are addressing a surface level issue. Just change your habit to be spaced repetition on flash cards.
wsl find ...
You can run all Linux commands this way. Also, pretty sure that find's "-o" is the Boolean "or", not "otherwise". (Yet another example of why learning from LLMs is dangerous, I suppose).
A naive interpretation of or in the light of Boolean algebra would be: do both and return true if either succeeds.
The very first question is full of obvious bugs.
You have 'find . -name "notes.txt"' selected, it then says 'You submitted: ls -R | grep notes.txt: find . -name "notes.txt"', then it responds:
'Thanks — your answer looks like it was partially entered.
'You picked find . -name "notes.txt" (good choice). The submission shows an escaped/unfinished string: find . -name \. The correct full command is find . -name "notes.txt", which searches recursively from the current directory for files or directories named exactly notes.txt.'
There seems to be some weird kind of quoting issue going on there. I would fix obvious issues like that before sharing this with the world.
You'll never make every single person on HN happy. But if you share your stuff early and make one person happy at the very least that means you should keep working on it!
Don't let perfectionism get in the way of good enough :]
Show HNs never pleased everyone, and it'd be silly to try, but until recently there was a bit of a "it's not perfect... but the person has spent more time working on this than I have even spent thinking on the problem" kind of expectation whereas now many of them feel like the comments section ends up doing more thought about the submission than was put into refining it.
It'll be really interesting to see where this settles. In the meantime, erring on the side of kindness tends to work best!
Anyhow, I'm just a bit bitter about the onslaught of obviously vibe coded projects that very little effort has been put into. And OP did ask for feedback. But I shouldn't have been so mean about it. I know it can be fun to be able to whip up something quicker than you could before, but I really wish people would spend a tiny bit more effort on it before asking for feedback.
This is far below most other "Show HN" posts and your first message was spot on.
There is a reason chat bots work - it mimics natural human interaction.
But this can be interest pattern - for say AI driven personal tutor for math topic
Lastly - for love of god - pls do something with the UI - all colors and bubbles and I am totally lost just trying to make sense of what is going on. Look at reditt if you just want back and forth thread convo style. Life is simpler that way.
Agree on colours. The Electronics topic was white on yellow, completely unreadable.
But if I have to read a wall of text, it's not an alternative to doomscrolling, it's an alternative to reading a book or documentation. And in that case I'd rather read a book or documentation.
This is something else entirely and requires far too much thinking for the label.
I am also not a big fan of trying to beat doomscrolling. One of the defining properties of doomscrolling is that it is mindless and addicting. The moment you try to create a mindful, healthy alternative, you've already lost. No product will ever beat doomscrolling, only individuals dedicated to their own mental health are capable of clearing this hurdle.
It would always start to make every correct answer option "C" over time, no matter what I tried. Eventually I was so focused on whether or not it was stuck in a "C" loop that I started overthinking all of the questions and wasting time.
Flash forward to testing Sonnet 4.6 recently to try and see if it could effectively teach me something new, I got about 5 prompts in before I had to point out an oversight, and it gave me the classic "you're absolutely right, ignore that suggestion".
This is anecdotal of course, but at least LLMs are helping to build my skills of fact verification and citation checking!
gripe: the "+New conversation" button is distracting, floating on the botton. what about floating a "+" in the upper right corner?
Your obvious first port of call IMO is correctness of material, where there's room for improvement [1]. I deliberately picked Gleam because it's still a less known language.
For what it's worth, prompting Opus 4.6 in chat got me this result [2]. Sonnet 4.6 via the Workbench also got it right.
Agree with other comments about UX and design, and maybe also some of those around improving teaching style or gamification aspects, but the above is more important.
Good luck with this, hope you crack it! :)
You upload a source and it generates questions from it. However when showing it to friends I found that the barrier to usage was too high as most people don’t have a source ready. But I think adding it as an option would be pretty cool and doable