sinaatalay

7 ms·

31.
by sinaatalay 3mo ago
No, not yet. It's something we may do in the future.
32.
by sinaatalay 3mo ago
We are. We're confident in the quality of the teaching and the accuracy of the content in these videos. Our videos are not some MP4 files sitting on a disk. We see our content like software. We review it, version-control it, and mainta
33.
by sinaatalay 3mo ago
We have an in-house computer graphics pipeline. We reimplemented 3Blue1Brown's Manim in Rust from scratch. It is available at https://studio.academa.ai . The renderer runs in the browser using WebGPU.
34.
by sinaatalay 3mo ago
By storing the content in a semantically meaningful way (not mp4s), making it easy to review, version control, and maintain.
35.
by sinaatalay 3mo ago
Calc 1 and Calc 2 will come later. We'll actually be doing this for hundreds of technical subjects over the next 6–12 months.
36.
by sinaatalay 3mo ago
In terms of UI, it's just videos and an LLM chat. Though, the videos include problems along with their solutions.
37.
by sinaatalay 3mo ago
The LLM has the video's full context, including visuals. Try chatting with it. Jump to any point in the video and ask questions about what you see.
38.
LLM-Integrated Multivariable Calculus Course(calculus.academa.ai)
42 points by sinaatalay 3mo ago|47 comments
39.
by sinaatalay 4mo ago
"Why is that? How could that be? The answer is because customers don't form their opinions on quality from marketing. They form their opinions on quality from their own experience with the products or services." - Steve Jobs
40.
by sinaatalay 5mo ago
This sounds to me like saying: "if poverty has decreased, everyone should be able to buy a 40m yacht." We can now simply buy an artificial form of intelligence to solve intellectual problems. That doesn't mean we can buy our
41.
by sinaatalay 5mo ago
If Photoshop can be vibe coded in a couple of weeks, that's superintelligence. I think of "tokens" as units of intelligence spent on a problem. Producing software like Photoshop, Linux, or Excel literally cost billions of tok
42.
by sinaatalay 6mo ago
It's not that AI is necessary, but it's that one may not choose to (or have the skills to) spend a whole weekend hand-coding a 3D interactive visual. But one might spin up Claude Code and build whatever the explanation actually ca
43.
by sinaatalay 6mo ago
Very cool to see these browser-native interactive 3D visualizations! Gives this such a different energy than a regular blog post would have had. I'm guessing those visualizations wouldn't be in this post if it weren't for AI.
44.
by sinaatalay 9mo ago
We want to nail the fundamentals first: Single-variable Calculus, Linear Algebra, Differential Equations, Physics, Machine Learning etc. These are the courses with the largest audiences, and we need that reach to sustain this project. But i
45.
by sinaatalay 9mo ago
Thank you! We actually made the course ourselves, following James Stewart's Calculus and our own approach to teaching these topics (we're two PhD students). We wrote everything in English first, then translated it into other langu
46.
Show HN: Open Computer-Animated Multivariable Calculus Course in 6 Languages(calculus.academa.ai)
5 points by sinaatalay 9mo ago|4 comments
47.
by sinaatalay 9mo ago
I couldn't find a link to the source code.
48.
by sinaatalay 9mo ago
I'm interested, but the "Watch demo" button doesn't work, so I can't tell whether this is what I'm looking for.
49.
by sinaatalay 10mo ago
You're absolutely right!
50.
by sinaatalay 10mo ago
Just to clarify, that commit was the result of about a month of careful development, and involved significant manual effort beyond AI coding. See the PR: https://github.com/rendercv/rendercv/pull/528
51.
by sinaatalay 10mo ago
Thank you so much for the kind words, and I'm sorry for the trouble the changes caused you. To give some context: I hadn't worked on RenderCV for about six months, and when I came back, I had grown technically and my design taste
52.
by sinaatalay 10mo ago
The JSON Resume schema did not support what we needed, so we deliberately created our own. We wanted to define 9 different entry schemas that could be used under any section title. In our schema, each section, regardless of its title, conta
53.
by sinaatalay 10mo ago
It’s completely flexible. However, it's not done through CSS. We use Typst to generate PDFs, and we template Typst files using Jinja2 from YAML. You can override the built-in Typst templates; see the guide here: https://docs
54.
by sinaatalay 10mo ago
Yes, unfortunately, you will need to convert them to YAML. I think an LLM can easily do that for you, though. This has been requested before, but we haven’t implemented it in order to keep our schema robust and simple. Maybe we should add i
55.
by sinaatalay 10mo ago
Thank you. We will add an analysis of how RenderCV PDFs perform with applicant tracking systems in docs at some point, but users have been reporting positive results. We generate the PDFs using Typst, which adheres to modern PDF standards.
56.
by sinaatalay 10mo ago
We do have that, but it’s a bit outdated. We’ll update it at some point: https://app.rendercv.com
57.
by sinaatalay 10mo ago
We deliberately chose not to use JSON Resume because we wanted greater flexibility. For example, in RenderCV, you can use any section title you want and place any of the 9 available entry types under any section. In contrast, JSON Resume ha
58.
by sinaatalay 10mo ago
RenderCV actually relies on its own Typst package: https://typst.app/universe/package/rendercv Here are a couple of reasons, just a few that come to mind, why using YAML + RenderCV can be more appealing than worki
59.
Show HN: RenderCV – Open-source CV/resume generator, YAML to PDF(github.com)
100 points by sinaatalay 10mo ago|41 comments
60.
RenderCV: A LaTeX-based CV/resume version-control and maintenance app(rendercv.com)
6 points by sinaatalay 2y ago|3 comments

More ›