danirod

The dubs for the Spanish version did it too. There were two dubs, one in 1990 and another one for a 2006 re-run. Both completely made up.

Sorry, am I going crazy or the text is LLM generated?

Is it intentional? Is it satire? Is it irony? Is it human generated but written as if it was written by Claude to be provocative?

I don't understand the comments section for this post, just like I did not understand the reactions when Codeberg did the same proposal a few weeks ago.

The kind of person who does the research and chooses to use Sourcehut or Codeberg vs just following the masses and using GitHub is exactly the kind of person who would know exactly what they're doing with computers and thus the kind of person that would not be afraid of just actually programming without the use of tools.

I don't have a Sourcehut account (nor a Codeberg account), but I follow a lot of projects that are hosted on those platforms and it's not just platform policies. The ethos is full of "we don't want that thing here". Lots of FOSS projects hosted on both platforms have their own "no AI" policy on the README or similar files.

Which is something that you would already know if you actually used any of these services.

To me, all these "oh my god, I can't believe they are not letting me host all this slop that I generated to fix my exact use case that is unique and will not be useful for anyone else" feels just performative drama.

AI;DR. Give us the prompt you used to generate the article, maybe it will be more understandable!

As someone whose inner (brain) chain of thought is messy, reading your prompt was a very pleasant experience. It might be dirty, it might not be polished, but reading this was like taking a dump of your thought, planting it on my brain, and letting me build upon it to understand your point.

I could not finish the output made by Claude. I tried, but having to translate back from "mouthy words" to "plain English" is tiresome.

I am not trying to say that I don't like well written words, a good book with carefully chosen words is a good way to spend an afternoon. But for LLMs, it feels very forced and not natural because it doesn't insert those words where it would make sense for a good author.

It seems that the link points to a shared map that is being flooded by visitors, so opening the link is a sensory nightmare. Plus probably triggers some bugs due to the server overload. Visiting the frontpage and creating an empty race result in a still buggy but less buggy experience to at least understand what in the world is this.

To be fair, I appreciate when they are so upfront about who wrote the code without requiring further heuristics, so I encourage this behavior.

With all due respect, but if humans can figure out how new unseen programs work by using -h and seeing what options exist and what they do, I am sure robots can figure it out too, or else they weren’t that intelligent to begin with.

It happened to me earlier today too. The response the machine gave me was completely different, as if it was a different codebase and a completely different task. It also had a few paths to some files and the username used for the `$HOME` directory was not mine. Local mode, not cloud or anything like that, and Auto mode.

Couple of "what the heck are you talking about"s, but because the attitude continued, I switched to Sonnet 4.5 and tried again, and this fixed the issue for me. I forgot to document it, and when I tried to export the original conversation for analysis it was gone.

But yeah, very confusing.

The Slopware paradox: I don't know if the error message is the actual website and this is just post-AI nihilism, or if it's a configuration error because if your website is not even a serverless edge function then `bro do you even scale`.

AIstor. They just slap the word AI anywhere these days.

Just give me the prompt, I'll seek the information by myself.