Live data from Hacker News

Launch HN: Miyagi (YC W25) turns YouTube videos into online, interactive courses

news.ycombinator.com

11–20 of 123 posts

Re: Launch HN: Miyagi (YC W25) turns YouTube videos into online, interactive courses

#12

How do you validate you’re not generating garbage, and thus teaching people nonsense?

For official courses, we go over the generated course with the creator to vet the content. Generally they're pretty impressed but have a few things they'd like to change/add before publishing.

For self-created courses, it's generally been quite accurate and we're playing around with some eval metrics to make it as good as possible, but it's definitely a concern.

Re: Launch HN: Miyagi (YC W25) turns YouTube videos into online, interactive courses

#13

Neat idea! Do you do anything with the video itself? Understand the visual content or extract details from slides?

Users can upload slides (ie. docx or pptx) and create a course from them - give it a try! For videos, we don't currently process any frames from the video just the transcript, but this is on the roadmap.

Re: Launch HN: Miyagi (YC W25) turns YouTube videos into online, interactive courses

#14
post #6

For anyone else interested in Bloom's 2-sigma, here's the original paper (1984): https://web.mit.edu/5.95/readings/bloom-two-sigma.pdf Blows my mind that 1:1 tutoring dwarfs the impact of other factors such as socioeconomic status, reinforcement, assigned homework, classroom morale, etc (at least according to the researchers). Does anyone know if this thesis has been replicated? Or if these results hold in modern tim…

Would be nice ie to see this product with focus on elementary school age content.

Re: Launch HN: Miyagi (YC W25) turns YouTube videos into online, interactive courses

#15
post #6

For anyone else interested in Bloom's 2-sigma, here's the original paper (1984): https://web.mit.edu/5.95/readings/bloom-two-sigma.pdf Blows my mind that 1:1 tutoring dwarfs the impact of other factors such as socioeconomic status, reinforcement, assigned homework, classroom morale, etc (at least according to the researchers). Does anyone know if this thesis has been replicated? Or if these results hold in modern tim…

The article states that Anaina and Burke separately conducted their tests, but social robots [1] have been shown to be effective in individual tutoring. Human tutoring is not always better than a well-designed computer program [2]. There have been issues with how studies interpret their effect on group size / scalability [3].

[1] https://www.science.org/doi/full/10.1126/scirobotics.aat5954 [2] https://www.tandfonline.com/doi/abs/10.1080/00461520.2011.61... [3] https://journals.sagepub.com/doi/abs/10.3102/0013189X2091279...

Re: Launch HN: Miyagi (YC W25) turns YouTube videos into online, interactive courses

#18
post #7

Are the people that create the content okay with this?

Who cares?

The people who spend hundreds of hours carefully creating content for their viewers care [0].

The referenced video is from a photographer who has some pretty strong and reasonable thoughts on this - specifically the features YouTube itself is experimenting with.

Depending on the nature of the AI product, it has the potential to completely sideline creators.

Not saying that’s what Miyagi is doing and it sounds like they’re actually working with creators on this which is good. But the broader point is that such tools need to be thoughtfully implemented.

- [0] https://m.youtube.com/watch?v=ULUSS1-G3do

Re: Launch HN: Miyagi (YC W25) turns YouTube videos into online, interactive courses

#19
post #10
post #7

Are the people that create the content okay with this?

Yes. Any content that we monetize we are revenue sharing with the creator. We already have more than 5 partnerships with creators.

3b1b is a monetized partner?

Association with that brand would be very valuable.

Re: Launch HN: Miyagi (YC W25) turns YouTube videos into online, interactive courses

#20

How do you validate you’re not generating garbage, and thus teaching people nonsense?

For official courses, we go over the generated course with the creator to vet the content. Generally they're pretty impressed but have a few things they'd like to change/add before publishing. For self-created courses, it's generally been quite accurate and we're playing around with some eval metrics to make it as good as possible, but it's definitely a concern.

Is the course creator being impressed the most important metric? Are there other more concrete metrics you are able to use to determine quality from the perspective of a student?

I am curious if you are using any methodologies from the digital learning space like knowledge tracing to help ensure that learners are actually retaining knowledge and improving over time or knowledge mapping to understand the gaps that might exist in your content?

Do you maintain your own skills taxonomy? Are you tagging your questions or assessment events with knowledge components or skills of any kind to understand what you are testing your students for?

All of this is really cool, I’m just curious at what level you’ve gotten to on some of this because there is a very fine line in online educational content between making the students life more difficult and actually helping them learn, especially when you get into auto-generating content, and especially if you aren’t following solid principles to verify your content. (I work for an online education company and particularly in the space of training LLMs and verifying their outputs for use in educational contexts)

Post reply on HN