[sixhobbits]

Some interesting stuff with a lot of LLM tells (part that actually matters, quietly, etc) which makes me skim instead of read.

Llm content can be good, but if I can't tell if it's slop or not at first glance (eg rent $800/year looked sus) I'm going to assume slop until someone I trust reads it and tells me it isnt.

[iJohnDoe]

Having a comment on every post about whether or not it’s written by AI needs need to be added to the HN Guidelines.

Read it. Don’t read it. AI written. Not AI written. Who cares. You don’t need to tell us.

[apothegm]

I can’t speak to the correctness of the numbers, but while it’s not particularly groundbreaking the analysis as a whole seems human and I enjoyed it. It’s got some tells that suggest LLM editing but is much more digestible than pure slop, at least.

[reasonableklout]

This particular article is reported as 100% AI-generated by pangram: https://www.pangram.com/history/52f51cff-9166-4b8e-ab2e-39fd...

I assume the author found the data sources and did some editing, and they have a few other articles that appear more human-written.

[bigbadfeline]

> This particular article is reported as 100% AI-generated by pangram

Your pangram is wrong, most likely deliberately so, there's material interest to train LLM-detecotrs to reject and taint everything that doesn't agree with the mainstream "alignment" constraints. This well written article is precisely such a case.

If you want to exonerate pangram, ask it to give you the prompt that generated the article. Or you supply that prompt yourself and tell us to which LLM to submit it.

It's a simple proof procedure, invented by myself, for the purpose of stopping the anti-llm SLOP which is being pushed on the web with a heavy hand. The offenders shall be shamed and remembered.