nestorD

5 ms·

1.
by nestorD 11d ago
I started programming at 18 right after highschool (which, I guess, is late by HN standards: a number of my peers had played with Python first and hated Ocaml).
2.
by nestorD 11d ago
I did so! (quirks of the path I took in the French educational system) I went through a book similar to the one above, with no internet connection. The first few weeks were rough: I did not quite know what a type was and the compiler error
3.
by nestorD 2mo ago
> For example there is less joy to be had in learning a programming language in 2026 than there was in 2016, because there is less utility in it now. I... disagree. The joy I get from learning a programming language does not come from kn
4.
by nestorD 2mo ago
Fun fact, unsafe does not let you turn off the borrow checker in Rust: https://steveklabnik.com/writing/you-can-t-turn-off-the-borr...
5.
by nestorD 3mo ago
I have wanted a JAX-like (XLA backed) Rust tensor library for ever! Plus it looks like they have support for dynamic shapes: https://tensor4all.org/tenferro-rs/design/dynamic-symbolic-s...
6.
by nestorD 4mo ago
See also jaxtyping which, contrary to what its name might imply, covers JAX/PyTorch/NumPy/MLX/TensorFlow arrays and tensors. https://docs.kidger.site/jaxtyping/
7.
by nestorD 5mo ago
It's what put MuJoCo on my radar recently! But I was surprised to not see him do any kind of gradient descent to optimize his hyperparameters. MuJoCo has a JAX backend so it should be fairly straightforward.
8.
by nestorD 6mo ago
For people wondering why the Islamic world would have had more texts, many of which are of western (Greek/Latin) origin, than the western world. The problem is that, as the Roman empire collapsed, papyrus supply disappeared in the west
9.
by nestorD 7mo ago
Her book Steering the Craft, is very much her writing workshop distilled into book form.
10.
by nestorD 8mo ago
In theory, yes! If this metric ever becomes a widely used standard, one would have to start accounting for that... But, in practice, when asking a model to pick the best answer they see a single question / answers pair and focus on det
11.
by nestorD 8mo ago
It presumes some models are better than others (and we do find that providing data with a wide mix of model strengths improves convergence) but it does not need to be one model, and it does not even need to be transitive.
12.
by nestorD 8mo ago
On alternative ways to measure LLM intelligence, we had good success with this: https://arxiv.org/abs/2509.23510 In short: start with a dataset of question and answer pairs, where each question has been answered by two
13.
by nestorD 9mo ago
I doubt it. The I Ching does not really have bad / low interest hexagrams. Also historians who studied the topic seem pretty sure that the yarrow stalk method is a recent introduction (by I Ching standards, we are talking about a bronz
14.
by nestorD 9mo ago
Fun fact: archaeological evidence on I Ching divinatory records shows an hexagram distribution different from the one produced by the yarrow stalk method. Meaning that, while it is now considered the traditional method, it was likely not th
15.
by nestorD 10mo ago
So far I have seen two genuinely good arguments for the use of MCPs: * They can encapsulate (API) credentials, keeping those out of reach of the model, * Contrary to APIs, they can change their interface whenever they want and with little c
16.
by nestorD 10mo ago
I started with a UI that sounded like it was built along the same lines as yours, which had the advantage of letting me enforce a pipeline and exhaustivity of search (I don't want the 10 most promising documents, I want all of them). B
17.
by nestorD 10mo ago
Oh! That's a nice use-case and not too far from stuff I have been playing with! (happily I do not have to deal with handwriting, just bad scans of older newspapers and texts) I can vouch for the fact that LLMs are great at searching in
18.
by nestorD 11mo ago
The paper[0] is actually about their logarithmic number system. Deep learning is given as an example, and their reference implementation is in PyTorch, but it is far from the only application. Anything involving a large number of multiplica
19.
by nestorD 1y ago
I have found putting the spec together with a model, having it to try find blindspots and write done the final take in clear and concise language, useful. A good next step is to have the model provide a detailed step by step plan to impleme
20.
by nestorD 1y ago
I have found Claude code to be significantly better, both in how good the model ends up being and in how polished it is. To the point that I do not drop down to Gemini CLI when I reach my Claude usage limit.
21.
by nestorD 1y ago
The first step is to acquire hardware fast enough to run one query quickly (and yes, for some model size you are looking at sharding the model and distributed runs). The next one is to batch request, improving GPU use significantly. Take a
22.
by nestorD 1y ago
One thing I could not find on a cursory read is how used were those developers to AI tools. I would expect someone using those regularly to benefit while someone who only played with them a couple of time would likely be slowed down as they
23.
by nestorD 1y ago
Yes! I recently had to manually answer and close a Github issue telling me I might have pushed an API key to github. No, "API_KEY=put-your-key-here;" is a placeholder and I should not have to waste time writing that.
24.
by nestorD 1y ago
I don't use it to avoid reading man pages. Rather, as often with LLMs, this is a faster way to do things I already know how to do. Looking at commands I run in various situations and typing them for me, faster than I can remember the n
25.
by nestorD 1y ago
Yes, it is API based and uses your last unique 100 shell commands as part of its prompt: it seemed important to remind users that this data does leave their machine. A fork using a local model should be fairly easy to set up.
26.
Show HN: Shelly, terminal assistant that translates natural language into shell(github.com)
13 points by nestorD 1y ago|4 comments
27.
by nestorD 1y ago
I think the top post on the Krita thread does a pretty good job at setting their boundaries. Something that cannot replace artists: it will not "beautify" art, and stays close to the input, also it should not be trained on the wor
28.
by nestorD 1y ago
Anthropics actually encourages using Claude to refine your prompts! I am not necessarily a fan because it has a bend towards longer prompts... which, I don't know if it is a coincidence that the Claude system promps are on the longer s
29.
by nestorD 1y ago
Nice! I will give it a try later today. For people who want a non web-based alternative, these days I use Xournal++ ( https://xournalpp.github.io/ ) to do that type of edition locally. What I am still looking for is a good wa
30.
by nestorD 1y ago
Yes! And the next articles in the series double down on this: "Any polynomial basis has a “natural domain” where its approximation properties are well-known. Raw features must be normalized to that domain. The natural domain of the Ber

More ›