stellalo

I’m wondering: could one achieve the same level of ergonomicity (lack of lifetimes, Rc, …) by imposing a similar “discipline” with thinking ahead about data storage, and staying within a subset of Rust?

Edit: I guess one would still miss Goose’s memory layout, which is (if I understand correctly) where it wins in performance.

> This is a pretty grim prognosis for European AI.

Is this better or worse than no news from European AI?

Prompt Repetition Improves Non-Reasoning LLMs: https://arxiv.org/abs/2512.14982

I don’t think you can tell someone why they are sad, better than they can, can you?

Was thinking the same

Does anybody have info about the battery life? Doesn't seem to be mentioned anywhere.

> Full auto-calendar (to year 2099)

This they will need to update pretty soon

harness + llm = agent

But the thesis of the OP is that it is !easier && different

You’ll fail the mandatory inspection

Oral exams, what an experience

This is really beautiful and tragic at the same time. Very well written.

I speculate the real goal is to have that fixed over time, and then use it as precious training data for Rust capabilities

Is this some sort of “incompleteness” paradox for AI alignment? Seriously

It doesn’t look like OP or the specific paragraph is describing an LLM problem, but rather a people problem

> A skill is a markdown file with frontmatter that gets injected into the agent’s context when the situation calls for it.

When the LLM decides that the situation calls for it

> It is a workflow: a sequence of steps the agent follows, with checkpoints that produce evidence, ending in a defined exit criterion.

A sequence of steps the LLM can decide to follow

I switched to Zed as my main editor about a year ago, and I’m not going back! What a great product!

probably: worked on it, then got sick of it

> I guess I will be trying the latest offering from OpenAI and Google tomorrow and if they are satisfactory I might just switch.

If Anthropic’s move is confirmed, my guess is other coding agents providers might end up making similar moves

One thing that is to be said about uv: you might come for the speed, but you stay because it works so well. The Python ecosystem used to be such a mess to manage, between Python versions, environments and so on: I haven’t complained once about it since I use uv. To the point that I tend to distrust a Python project if it’s not managed with uv, and anticipate all the dependency resolution issues that I will face.

This should not be flagged.

Flagging this: that’s fascism.

Doesn’t Claude Code allow to just dump entire conversations, with everything that happened in them?

I was also disappointed by the lack of Jupyter notebooks support: I ended up not using Jupyter notebooks that much anymore, and when I do, well, I run them in Jupyter

This is the same reason I switched to it. It’s just lightweight and fast, I don’t see why a text editor should not be like this

For me, Safari sometimes randomly refuses to execute the search for the terms I entered: at that point I need to bring the search bar back up -> search terms are gone -> x -> bring search bar back up -> search terms are back there -> enter

I wish they stopped adding features, especially useless UI “improvements” and AI stuff nobody asks for, and focused on making the system rock solid as we’re used to.

you also need check_if_negative to detect close-before-open

I feel like it’s the opposite: the copy-paste issue is solvable, you just need to equip the model with the right tools and make sure they are trained on tasks where that’s unambiguously the right thing to do (for example, cases were copying code “by hand” would be extremely error prone -> leads to lower reward on average).

On the other hand, teaching the model to be unsure and ask questions, requires the training loop to break and bring a human input in, which appears more difficult to scale.

> AI slop is mostly noise. It doesn't manipulate

Not until you start mass-producing fake photos, fake videos, fake audios, put all of it into social media, shake shake shake.

> The extra typing clarification in python makes the code harder to read

It’s funny, because for me is quite the opposite: I find myself reading Python more easily when there are type annotations.

One caveat might be: for that to happen, I need to know that type checking is also in place, or else my brain dismissed annotations in that they could just be noise.

I guess this is why in Julia or Rust or C you have this stronger feeling that types are looking after you.

Title should be fixed: it’s about Claude models in general, not Claude Code