cjv

Congratulations Stefan, "Kaggle for robotics" seems like an exciting direction! I still remember your kindness and just wanted to wish you the best.

I wanted to 3D print infinite LEGOs, but I was told that typical machines are not precise enough for this task.

I'm curious to know what people think about elisp as a extension language. Would Emacs be better if everything was implemented in elisp including the core data structures (assuming elisp was very fast)? Do people suffer from lack of static typing?

Wow, that is a beautiful animation at the end there.

Reflect is a cool name, but nearly impossible to Google.

Can you name some breaking changes PyTorch made?

Eigenvectors for building a leaderboard? Can you explain more?

Kubernetes did it first.

How does APL compare to J? Which one is more enriching to learn?

Ah, thanks for the explanation.

...doesn't the JAX example just need the argument set to static_argnums and then it will work?

PG seems to agree with Amjad's decisions.

https://imgur.com/a/WkpHk0n

https://twitter.com/amasad/status/1401617251464138754

Unbelievable and shameless.

Very ignorant comment

https://en.wikipedia.org/wiki/Einstein_field_equations

https://en.wikipedia.org/wiki/Lewis_Carroll#Mathematical_wor...

https://en.wikipedia.org/wiki/Ada_Lovelace#First_computer_pr...

How does this relate to a Fenwick tree?

(Edit: The article did not include its note on Fenwick trees when I commented this.)

I'm always impressed by SpaCy's web design, it's so distinctive and the documentation is a pleasure to navigate and read.

Chez Scheme is a good example on how many compiler passes can be extremely effective.

https://github.com/nanopass/nanopass-framework-scheme/

I've always been curious on what Mike would do if he could completely rewrite D3. I saw some experiments with a declarative API [0] which seems really interesting.

It seems like the main learning curve for learning D3 is deeply understanding the underlying web technologies (looking at you SVG), so could something D3-like have a more abstracted renderer?

Anyway, once you use Observable - Jupyter notebooks are forever ruined, which is the highest praise I can give. Thanks for all your hard work Mike!

[0] https://observablehq.com/@unkleho/introducing-d3-render-trul...

Can you give a situation where CTEs are still a optimization barrier in Postgres?

"The Klutz Book of Paper Airplanes" is one of my favorite books from my childhood, having (by far) the best paper airplane designs. Thanks for the many hours of fun!

How did the research process work while developing a book?

No, I'm not wrong.

Detectron was open sourced in 2018. R-CNN didn't have any public implementations (there was later a Keras implementation that didn't get the same performance as the paper reported). TensorFlow models added some object detection models a few days after I started my internship, but had various issues at the time. SSD and YOLO both had public implementations, YOLO's being in it's own C based framework.

It's a completely different landscape three years later.

This is sad news.

I worked at Starsky Robotics as a perception team intern after graduating high school. I will always be grateful for the team for the opportunity, it was a fantastic first job and everyone who worked there was very kind (especially Stefan).

Unfortunately, Starsky had effectively had no machine learning in 2017 (when I worked there), using solely classical computer vision techniques. This didn't match the company's ambitions of not using LIDAR and there was a strong stigma against switching to a deep learning approach. At the time, very few object detection models had public implementations and I spent a lot of time trying to get a YOLO9000 and RetinaNet implementations running at real-time speeds. Frustrating, as a small startup the labeling services kept screwing us over by returning poorly annotated images.

I think what I took away from the experience is that deep learning in domains with long tails requires a enormous investment in a labeling pipeline - dwarfing the computational aspect - to get decent results. I don't think any solutions are on the horizon that will allow us to bypass this reality. You don't see improvements between Comma.ai and Tesla because it's about the improvements far out on the tail.