Hacker News Reader

Comment by netsharc | Hacker News Reader

What an oddly written article. ChatGPT? SEO bot? High-schooler?

These sentences don't seem to have a very strong connection with each other:

> Getting black currants banned has been deemed minimally effective for disease prevention. Therefore, some states start reversing the federal ban on this berry. However, Europe still remains the producer of 99% of the world’s black currants stock.

"It is not so bad"?

> Today, these plants are successfully grown in New York, Connecticut, Oregon, and Vermont. Yet, the majority of Americans can only enjoy processed or dried berries. It’s not so bad considering the benefits of eating dry fruits.

View on HN

It's badly written, but more on the level of a bad human writer than any AI. The connection between the three sentences in the first example seems quite clear to me. Starting from the second sentence:

1) "[...] some states start reversing the federal ban on this berry [because the ban has been deemed ineffective for disease prevention]".

2) "[Even though some US states are reversing the ban] Europe still remains the producer of 99% of the world’s black currants stock".

Regarding the second example, it is just an excuse to have an internal link to another article on the same site praising dried fruits. "It's not so bad [that Americans can only enjoy processed or dried blackcurrants] considering the benefits of eating dry fruits.". The reasoning is quite crappy, given that the linked article does not imply that dried fruits are better than non-dried fruits, but it is to the level of precision that one might expect from a filler blog on some health-food online store.

The publishing date of the article is 2017, so pre-ChatGPT. Of course there were spintaxes back then, but those articles were way worse.

Either way, the article seems mostly fine to me.

> The publishing date of the article is 2017

I used to predate my blog posts in Wordpress by years. Why would you use todays date? No one reads blogs in order, how pedestrian to follow the old rules.

You can also fool some of Googles date metadata, it's not just the day it's first indexed. Reddit by incompetence screws with it.

But I've never seen a content farm do it.... yet.

Webarchive first snapshot 8th Nov 2020 - http://web.archive.org/web/20230000000000*/https://foodtoliv...

The idea there is a pre-AI internet is not true.

AI will go back and change the past. Not even Google/Web Archive can record it all. Is it AI pre-dating or was it missed on Web Archive here? (A: The Facebook comments seem real, so missed by Web Archive, but if Facebook deletes/makes private the data then we are back to not knowing)

Chatgpt wouldn’t write crap like that

Sounds like an ESL writer, 'It is not so bad' is a common phrase I see written by native slavic speakers, it's probably a direct translation of a language idiom.

Why are we talking about it uncontracted, making it sound weird, when the actual quote is 'it's not so bad' which sounds perfectly fluent to me (native BrEng)?

It’s definitely a common phrase but it’s really poor writing here.

I agree, just only really because it's too colloquial/informal/spoken for an article imo, wouldn't be my choice. But not worth calling out, and certainly not for not making sense or seeming like a robot or ESL author wrote it.

It also answers the question directly in the first paragraph with no fussing around, so I like it better than 99% of articles.

It reads like it was automatically translated from another language. The grammar is all over the place.

the weird aff link in the middle for random berries is the cherry on the top

Definitely reads like AI. Short declarative statements. No coherent narrative.

(I see what I did there (o;)

Reply on news.ycombinator.com