Top all-time · #7208
Patrick Kidger
Personal blog · kidger.site/thoughtsaiprogramming
Patrick Kidger's personal blog at kidger.site covers ai and programming. Since 2022, 11 different Hacker News readers have linked it 14 times, most often “Jax vs. Julia (Vs PyTorch)”. It is still writing: the last post was on 2025-05-20, and you can follow it by feed.
Too few sampled comments (7) to say how often HN recommends it.?
Why HN links it
HN readers link Patrick Kidger's JAX vs Julia post as a reference on Julia's autodiff reliability and ecosystem quality compared with JAX. Commenters cite it in deep learning framework discussions.
Summarised by Claude from a sample of 7 HN comments that link here. Each comment is marked recommending, critical or neutral (citing without judging); how this works.
I do scientific machine learning: antibody therapeutics, neural ODEs, open source software.
Most-linked posts?
- Jax vs. Julia (Vs PyTorch)4 people · 4 mentions · first linked 2022-05-04
- Just know stuff. (Or, how to achieve success in a machine learning PhD)3 people · 3 mentions · first linked 2023-01-27
- /2 people · 2 mentions · first linked 2023-08-14
What Hacker News says?
This is an old-ish article about Julia, but from what I can tell the core issues with autograd were never fixed: https://kidger.site/thoughts/jax-vs-julia/
Comment linking /thoughts/jax-vs-julia/
on “Micrograd.jl”Submitted Jax vs. Julia (Vs PyTorch) as a story · 141 points
HN discussionhttps://kidger.site/about/
Comment linking /about/
on “How to solve computational science problems with AI: PINNs”Submitted Just know stuff. (Or, how to achieve success in a machine learning PhD) as a story · 27 points
HN discussionThere's lots of good reasons. Most notably the autodiff in Julia is unreliable -- there's been quite the litany of packages trying/failing to do this robustly and efficiently. (This issue pretty much kills its possibilities straight from the get-go.) The fact that the ecosystem is largely created by academics is also a serious issue. Code quality is very low / things silently break / etc. This is partly an issue wit…
Comment linking /thoughts/jax-vs-julia/
on “Ask HN: Why hasn't the deep learning community embraced Julia yet?”Submitted Just know stuff (or, how to achieve success in a machine learning PhD) (2023) as a story · 3 points
HN discussionSoftware engineering is hard problem. How is Julia in multiple authors codebase? A language need to consider ergonomics of multiple users working on same thing. Reading >> Writing. C++/Python still hold ground in this regards sadly. One side for low-level control, another side for glue code. Related: https://kidger.site/thoughts/jax-vs-julia/
Comment linking /thoughts/jax-vs-julia/
on “ML would have advanced faster if another front-end language had been adopted?”Submitted Just know stuff. (Or, how to achieve success in a machine learning PhD) (2023) as a story · 3 points
HN discussionYes, I do remember a few people once I read their handles, like this guy [1]. [1] https://kidger.site
Comment linking /
on “Show HN: Anonymize HN – a Chrome extension to anonymize Hacker News users”Submitted 'Real' Semantic Versioning as a story · 2 points
HN discussionThe Tiny Pharma Company | Founding Scientist | London/Zurich https://tinypharma.bio/jobs/ AI+pharma startup, and each person owns a program, from concept to clinic. (Further details omitted, still in stealth.) We hire anyone exceptional in AIxBio, we don't look to fill defined roles. Founding team (including me! https://kidger.site ) are ex-{Cradle, Google, Oxford, ETH}, backgrounds in protein language models, antib…
Comment linking /
on “Ask HN: Who is hiring? (July 2026)”Submitted Just know stuff, proteinML edition as a story · 1 points
HN discussionThat's right, but at the same time Flux, the default choice is not mature at all compared to JAX. So that advantage is only on paper. I would really prefer to use Julia, I actually dislike Python, but it's unrealistic to implement large architectures in Flux. It is buggy, lots of gradient calculations are unreliable. See some discussion here for instance: https://kidger.site/thoughts/jax-vs-julia
Comment linking /thoughts/jax-vs-julia
on “Makie, a modern and fast plotting library for Julia”Submitted How to handle a hands-off supervisor as a story · 1 points
HN discussion