Hacker News new | past | comments | ask | show | jobs | submit | Valk3_'s comments login

I'm kind of late to this post, but really awesome initiative and well executed project! Thank you for bringing this to people!

I tested it a bit and it seems pretty decent, although for some really niche theoretical questions it wasn't successful in retrieving the answers I wanted even if alot of the results were really good in other aspects. It could simply be because the answer is not available anywhere in hackernews.

I'm wondering if someone were to build a similar project but for other sites, what would your advice be? For instance what technical difficulties did you stumble on that you think would be good to be aware of?

Thanks in advance and once again congratulations on the project!


Thanks for the kind words!

Yes, the underlying dataset very much conditions the quality of the responses. Additionally, the retrieval strategy is also a really important factor (and that is something which I haven't had time to extensively optimize).

I'm writing a blog post that will answer your questions! Will post it here when it's fully baked.


Awesome, looking forward to it!



Great work! I got a question though, what's the processing time for longer videos? Like lecture videos that usually are somewhere around 1h-2h? Can it handle longer vids reasonably fast?


Should be in the ballpark of 30s to 1 minute. You should start seeing earlier parts of the summary faster than that though (< 10s). I am also working on improving chapter summary quality for long videos - sometimes too much content gets rolled into a single chapter.


Another alternative to hanging out in forums is to join a good math discord server. I can highly recommend this one: https://discordnetwork.com/, I have no idea how many people are in the server but I'm guessing at least more than 10k (about 9k online as I'm writing this, and more offline I have to presume). The advantage with such servers is fluid communication, but the downside is because of this in combination with the fact that there might be alot of people, you sometimes have to wait as to not disturb ongoing discussions.

This discord server in particular is good if you want help, although you might not recieve instant help, you will eventually get help if you are patient. There are also other channels in the server just for the purpose of discussion, i.e discussing general topics in math.

This might not give you what you are looking for, but it's a possible alternative I guess.


I don't see the technical aspects of this project being discussed as much. Could you the author share some of the technical difficulties with a project like this? For instance, what were the hardest part to make this work? What are some improvements that you are planning on doing? Since it seems that the planner still has room for improvements as can be seen in the discussions here.


What would you say are your biggest take-aways from that course?


Here are a few:

* Diffuse mode vs. focus mode. After focusing hard on a problem, letting your brain wander can do wonders for coming up with insights and ideas. The classic example is coming up with something in the shower after working on it throughout the day. I've focused much more on giving myself some of the non-focused time after focused periods (ex. going for a walk/run/swim, taking a nap or shower, etc). I've started doing this more for work, as well.

* How memory works (short-term vs. long-term) and along those lines, spaced repetition. All through undergrad I would cram, but spacing it out (with the help of Anki for flashcard-focused topics) really does wonders.

* Importance of actively quizzing yourself, practice, and working through problems as you're learning something.


I’d recommend asking Tim Ferris.


Mr. Ferris certainly has a lot to say about the process of learning, but I haven't run across him saying anything about this particular course.


This was a joke.

Tim Ferris still promotes a “life hack” of paying other people to read books for you and provide a summary. Of course, this entirely detracts from the entire reason to read a book or any complex piece of information that might affect people differently depending on biases and past experience (so anyone who’s interested enough to read said information / content).


Dunno if people find this useful, but there's a program called cold turkey that blocks certain sites that you can specify for a period of time that you also specify.

I think it only supports Windows and macOS though. https://getcoldturkey.com/


Just curious, what are better alternatives to desktop apps? And why is python + Qt bad?


Agreed, would also like to hear more about this!


I'm surprised nobody has mentioned Dzone yet.

https://dzone.com/


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: