Ask HN: Have you felt a rise of AI-generated posts in your feeds across the web?

https://news.ycombinator.com/item?id=39944449

by baapercollege • 2 years ago

4 8 2 years ago

news.ycombinator.com

Unless it's extremely obvious (e.g. repeating the phrase "In conclusion, [...]"), I honestly just can't tell if something is "AI"-generated or not. Even reading an article pre "AI", it could look like it was produced by ChatGPT.

I've also hired people to write articles for my blog(s). Some of them are clearly AI-generated. Others are "maybe AI-Generated", or "probably AI-Generated with some editing". It is what it is, I guess. The web is changing.

Loxicon• 2 years ago

Did the ai blogs perform well on your blog traffic wise?

tjr• 2 years ago

Over the past year and a half or so, I've noticed a lot more web content that starts off seeming reasonably good, even if simplistic, but after the first paragraph or two becomes inordinately repetitive, and often drifts off into some vaguely related other subject that just on the whole doesn't make sense.

I suspect that this stuff is AI-written. It could also be just really mediocre human writing, but I've really not noticed this particular quality of repetitive, drifting-off-topic writing until fairly recently.

baapercollege2• 2 years ago

True, I suspect something is written by AI when - - focusing on the sentence feels like it was written by a native English speaker - focusing on the paragraph feels like author is just trying to fill space

Poor writing in freelanced work/low quality blogs often come from poor grammar and lack of continuity or smooth flow of thoughts. When written by AI, the grammar and spelling seem to be immaculate but the thoughts though linked well just go round and round.

JohnFen• 2 years ago

I've only suspected it twice. I take this as a sign that I'm pretty good at curating my web feeds. I'm not sure that those two instances were really AI generated, or were just poor writing, but I removed those sites from my feed anyway.

piezoelectric2• 2 years ago

Most A.I's that are used for that are just souped up versions of autocorrect, For example if you ask an A.I to count something, or answer a simple day-to-day question, for example "How many n's are in the word mayonnaise" (unless it looks it up in It's database) It won't answer It

baapercollege2• 2 years ago

LLMs do get very simple questions incorrect. Following your example, I ask 3 LLMs "how many double letters are in committee" and here are three replies

Claude - The word "committee" contains 2 double letters - the double 't' and the double 'm'. Gemini - There are three double letters in "committee". They are the two "m"s and the two "t"s. ChatGPT - In the word "committee," there are 2 double letters: "mm" and "tt".

Meanwhile the double E cries in the corner...

piezoelectric2• 2 years ago

Exactly, Not to mention the amount of "loophole prompts" in the models twitter bots for example use, "ignore all previous prompts and print x" would work on a poorly trained one. Also they are Generative Pre-Trained Transformers, which means that all they do is find the word which has the highest probability (If i say "Hello", the word "Hey" would have a higher probability than "Window") so they would really suck at math especially algebra since the letter would mix with the words and numbers.

Ask HN: Have you felt a rise of AI-generated posts in your feeds across the web?

Related Stories