Hacker News Reader

Comment by theletterf | Hacker News Reader

They had the guts and willpower to learn Sumerian. Why, I wonder, had they to use an LLM to write the essay about it? It kind of invalidates the core messages.

View on HN

I read the entire post, very well written, and didn't suffer from that impression that it was written by an LLM

It uses words. A lot of words. And punctuation.

And spacing. To say what?

Very little.

Yeah it's so sparse on actual details of Sumerian or cuneiform that after reading the article I remain unconvinced the "author" either cares for or knows anything about the topic.

(To be clear I know nothing about it either, but as a human with some similarly "pointless" hobbies, I can't imagine posting a writeup with so much motivational bullshit but so little about the actual thing I'm interested in.)

You are an engineer who's recent project was cancelled due to a change in priorities. It left you feeling burnt out but eventually you sought solace in learning Sumerian, just for the sake of it. Please write 1500 words in the style of a motivational linkedin post.

The entire "Why Irrelevance Works" uses several LLM patterns. As I said elsewhere, I don't care if someone uses GPT to coauthor a post, but the irony is just too big in this case.

They work in tech, they have a tech blog with articles about AI, so it's reasonable to assume they use an LLM to generate their content.

I'm willing to give them the benefit of the doubt that they ever actually tried learning Sumerian, but to be maximally cynical they never provide any specific personal details about that. It could literally just be a prompt.

GPTZero says its 44% AI generated

GPTZero also says one of your previous comments is partially AI generated though.

Those models are terrible. I tried one of the "best" ones and it told me the US Constitution was AI generated.

Maybe it was?

(I am reminded of a Fallout side quest, where you help a Museum retrieve artefacts and they talk about how the constitutions was moved around, maybe by airplane .. point being, history will be fun to reconstruct after a robot nuclear apocalypse)

What makes you say that? Sad how easy it is for someone to dismiss writing these days as, "must be AI written!" (I'd wager the average LLM blog post takes more effort than your dismissive response, though I wish I didn't have to read either).

I'm not a fan of the author's writing, but you can look at their other articles and see the same (non-AI feeling) style of writing and general theme of content (airport-book style motivation).

I usually don't care too much if a post was written using AI (this one was written at least partially using it, check it out using GPT zero), but in this case it irks me deeply, because the post is meant to summarize deep intellectual effort.

I find it hard to believe someone who is clearly used to deep work would take a shortcut like that. I'm not saying they didn't, just that it doesn't fit.

Reply on news.ycombinator.com