semilin

You'd have to be incredibly ignorant to suppose it to be a work not worth reading. It is not the best piece of writing, not even of Marx's, but its influence on humanity is enormous; that should suffice for the value in reading it.

The French is pretty much word-for-word the same as the English, so the Claudeisms didn't come from translation. It's all the same LLM output.

"Cookery simulates the disguise of medicine, and pretends to know what food is the best for the body; and if the physician and the cook had to enter into a competition in which children were the judges, or men who had no more sense than children, as to which of them best understands the goodness or badness of food, the physician would be starved to death" (Plato, Gorgias).

Anyone who believes the objective quality of the end result is the sole end goal of all tasks will have no self worth when technology surpasses them. This is true for essentially any task, because there is almost no end result that cannot be done "better" by technology.

However, the blog post explains that this is the wrong way to think about writing. The very first point made is that writing makes you understand your subject better and develop better ideas.

Characterizing the argument as 'thinking oneself to be Shakespeare' is a silly strawman and helps nobody.

For small projects and hobby-programming, OpenCode Go is great, and its model performance is quite strong in my experience. Every time it's mentioned, there are people loudly claiming that it has terrible, quantized models, though this is never backed with data. I'm suspicious that this is being propagated by those whose financial interests are harmed by the existence of a cheap and decent coding subscription.

Looking the same is not itself the problem. This is a same-looking that indicates a lack of human thought underlying it. That is a problem, and it is concerning to me if one does not think this.

What a rude approach! I will always prefer the more classic and civilized way of waiting for the dictionary to invent new words for me to use.

Why? LLMs are not humans.

Or you could have read it more closely before posting a comment saying it didn't make sense. You know, the old school way.

What does this have to do with Vim?

They can be in the training set but not deliberately trained for. There may be a lot of people posting pelican svgs, but not typically because they're high quality and worth replicating.

Absolutely. While we're at it, why should asbestos be regulated? You can just not go in buildings that have it.

"Nuanced discussion" doesn't necessarily mean the sort one would have with a human. Statistical apologies are never going to be meaningful. One could edit nonsense into the context window and the model would attempt to rationalize it. The models are smart but you need to use them in a way that makes sense for what they are.

This is entirely irrelevant. The problem is that anybody can make a Microsoft Form and there's no obvious indication that it's a form from Github without digging around.

Those are just hyphens, actually. En dashes are for ranges (e.g., 1–4, which is admittedly hard to distinguish from 1-4), not compound words. Point stands though—LLMs do love compound words and dashes in general.