> I understand that Microsoft is planning to remove most of its Claude Code licenses and push many of its developers to use Copilot CLI instead. While Claude Code has been a popular addition, it has also undermined Microsoft’s new GitHub Copilot CLI coding tool — a command line version of GitHub Copilot that runs outside of development apps like Visual Studio Code.

And people here are interpreting this as related mainly to the Claude burning too much tokens too quickly and suggesting Microsoft should rather use SomeOtherLLM©?

Is this Hacker News or rather Marketing Wars?

Makes me wonder if this AI flood uncovers the unflattering truth about this community acuteness, or it's only a failure of existing guardrails and we just need to change them.

We learn simultaneously that 'your team' shouldn't rely on any one of those patterns ('none of them is enough'), but that pattern 1 'alone will surface a useful amount of fraud'.

We also read strange sentences like "Every analyst on your team will use them (ie window functions) once they exist, and adding the next fraud pattern stops being a project. [end of paragraph]"

Or irrelevant discussions about how filtering by "IS NULL" might be not applicable when almost none of the provided examples uses it (and the one which does uses it in different context).

This is low quality and too long.

My general location is also wrong.

This site's theme is barely visible.

And the entire idea for the site is at least couple decades old.

Unoriginal slop.

Yes, exactly this.

If I notice it means your PhD-level 2027 ASI technology failed. Since when HN is a place for boasting about failed projects?

I don't agree with you. They are not at all like the examples you mentioned. Calling something "AI slop" signals that the writing either fails to raise any important point or, even when it raises a decent point, it is so repetitive it wastes time of readers. This is not only a style problem.

To put it in LLMish: It's not 'tangential to the content' – it's directly addressing (the lack of) the content.

If LLM worked perfectly we shouldn't have noticed the text was generated. I and others did. I feel it's important to point it out, if we don't want low-quality texts fully flooding us.

> By the end of 2026, 90% of the articles here are probably going to be AI slop, and it will be totally useless to complain about each and every one of them.

Using the policy you personally adopted it surely will be so. I don't think news aggregator comprised of junk information is something which should be embraced, so maybe reconsider your position?

Yes, I know, but what motivated me to ask was that from my observations also less derisive comments raising AI point are prone to being downvoted. Like this comment I linked to was 'just asking a question'. And I saw others being more pleasant, with no different results.

Usually the LLM generated texts they are reacting to aren't IMO worthwhile - like in this case. Idk, I feel very surprised by how accepting of them others here seem to be (if measured by points system).

A lovely bit of AI slop.

Edit - This is not the first time I'm observing this. Could somebody explain to me why the comments which point out the discussed texts are AI generated are being frequently downvoted on Hacker News?

In the very same thread there is this apparently downvoted (as of now) comment: https://news.ycombinator.com/item?id=47807528

Why is it so, is this really this community's stance on LLM-generated, mostly weak and empty writing?

IMO both are simply bad and both contain usual telltales in spades (continuity problems, failed or trite metaphors/analogies, semantic failures, overall feeling of 'wtf is even being attempted here').

I'm not so surprised that people struggled with identifying [1] as human - the confounding factor is that this flash story is unpleasantly written, and it's not easy to realize that its failure modes (eg. trying to cram too much in too short a text) are rather human like. And I'm sure the fact that arguably the hardest to digest and rather bad human story opens the poll might somewhat influence the further analyses.

As others in the poll I failed to identify [5] as AI even though in hindsight the telltales are also there. That's because I rather liked it, and as a result it was harder to be vigilant. I also was very undecided on [8]. Finally I scored 6/8, but I wouldn't say it was easy.

Shame that comparing to the previous contest https://mark---lawrence.blogspot.com/2023/09/so-is-ai-writin... is not straightforward. In that one I scored 9/10 while having very easy time (I didn't even finish reading some of them before making up my mind). I also felt completely excused with my only failure, incorrectly identifying as AI the story written in the style of exhaustingly banal fan fiction. But frankly I found almost all the human stories in the previous edition better then the current ones.

In retrospect ChatGPT4 was a terrible writer. ChatGPT5 seems to be an improvement to the admittedly worrying point. Still not impossible to discover though.

However these are my impressions only and it looks maybe I was lucky and I should not generalize it? According to the website people had serious trouble discerning gpt4 writing also 2 years ago. And I'm rather shocked they did. And that they scored some of those banal AI stories positively.

If it's not luck on my part, then maybe discerning AI writing is a skill very different from 'writing' or 'being deeply interested in literature', skills of people who usually frequent this blog?

What's the point?

The European Dark Ages narrative was indeed overblown and needed correction. But this correction went too far. It seems to be now at the stage of explicit and vigorous denial of any downfall of fortune in the Western ex-Roman provinces. I'd posit that such a denial is even more overblown than the initial myth it aimed to correct.

I can offer you a link to an author arguing for this position: https://slatestarcodex.com/2017/10/15/were-there-dark-ages/