mrborgen

Thanks. Definitely lots of work left to make this feel non-AI-slop.

Think there's a ton of potential here too. So we're staying on course.

Hahah! Thanks.

We've always gotten a ton of positive comments from people with ADHD who've used our HTML videos to learn to code. Despite it not being in our target group at all when we started out.

Hope your friend finds it helpful!

Thank you!

That was one of the challenges when building this. We ended up using Firecrawl. But even so there are a few pages we aren't able to read. In which case we use the Algolia API to get the HN comments and generate the video about the HN reaction on the article instead.

We currently spend ~$0.4 on a video (without images) and a have a few seconds of delay before playback. So it's a matter of turning every stone in order to optimize it further. We're going to add Jev asap, as it's an obvious improvement along both axes.

We usually include a section about the comments in the end of the video. But perhaps we should add a dedicated "Comments Video" as well.

Glad to hear that. Adding Jev as our "pre-planner" is very high on our todo list. It's the perfect tool for this kind of thing.

We do cache them like that! As I wrote:

"We create them on-the-fly the first time someone clicks on a link."

Though I realise it could have been said more explicitly. But the second time someone clicks the link we re-use the same explainer.

Btw, we are hiring a new developer to join us in Oslo, Norway. We pay well and give a lot of freedom wrt working from home or at the office (hybrid solution). If getting HTML-based video generation down to 1 cent for 1 minute with 1 second delay, then please email me: [email protected].

It doesn't actually capture it as a video, it's just HTML with a voice over essentially.

Nice! Will check out that repo.

You can use MCP to generate an Explainer video next time if you want. It's currently free, as the tokens for directing and scripting it is offloaded to your agent. And then we cover the TTS for the time being.

I hope so! We actually already support vertical videos when generated via the web ui (by clicking the mobile button).

We are not particularly pleased with the state of our visualizations, so I totally agree. If you have any examples of how you'd like to see the visualizations, please share them!

We have a free MCP so you are free too use our tool if you'd like to use it in your courses as well :)

Yes! We can set that up as a video quite easily. Don't have the infrastructure to push it as a podcast yet though.

Thanks! And I agree. We're getting tons of requests for more nuance in voice selection from our users. You're currently able to say i.e. "Australian English female" in your prompts today, but you should also ideally be able to describe the voice characteristic (i.e. like a funny grandpa, engaged news reporter).

What Gemini 3.8 Flash TTS is doing with generative voice design in this area super interesting.

That video in the README is amazing! Will definitely dig into this project.

We're not using Opus 5.5 btw, as it would be too expensive. Our goals is to eventually get the cost down to 1 cent for a 1 minute video, and with 1 second delay from submit to playback. For that, we need dirt cheap and lightning fast models.

Thanks! Several people on our team are actually exactly like you: they don't really use videos themselves for learning, but see the utility for others, and enjoy the technical challenges around building a high-performant video format.

Here is a Loom on how to record a scrim: https://www.loom.com/share/1f034653a1ef4e99953cf3b9a5c367ff

Here is a quick Loom on how to create a scrim: https://www.loom.com/share/1f034653a1ef4e99953cf3b9a5c367ff

Yes, you can! I'll record a video on how to do that with the new IDE, as it's indeed a bit unclear.

Technically, there are ways to do that, though it's not a core use-case for us, which is why we're not pushing it.

An alternative way is to embed the scrims into your own platform, and have the scrims be unlisted on Scrimba. Would that do?

Just send me an email at [email protected] :)

Once we have our desktop app ready (very soon), it'll be possible to record any language. However, the viewer then won't get the same ability to run the code, only to edit it.

We do have rudimentary Python support with the interactivity though. Planning out a course on it these days.

Yes, we'd be happy to collaborate on creating something on Joystick. It looks like an awesome framework!

We currently create the courses ourselves, but would love to see if there’s an opportunity for a collab here. Please send me an email at [email protected] :)