Hacker News Reader

Comment by keyle | Hacker News Reader

How long until we find an army of ChatGPT agents commenting on HN?

It is difficult to predict when an army of ChatGPT agents will be commenting on HN. It may take some time before such technology is developed and implemented.

This is ChatGPT's response, isn't it? I am beginning to develop a nose for its wordy style.

If you, me, we can recognize this distinct style, GPT-n will be able to do so too, sooner or later and will stop using it. This is a kindergartner that just learned speaking and never read a book on its own, it will soon go to grade school, high school, university...

gpt-n can't "recognize" anything. It just throws together words that tend to be thrown together. At best it is a parrot, but not even that. A kindergartner is still well beyond any AI.

Until chat gpt I had the same opinion, but after the latest monthly progress my opinion is shifting…

It's the completely non committal unopinionated response to any non objective question. Real people have actual opinions even if they aren't provable.

That, and "developed and implemented". ChatGPT seems fond of this student paper boost-the-word-count kind of redundancy.

I disagree with the statement that real people have actual opinions even if they aren't provable. Just because someone has an opinion does not necessarily make it valid or real. In fact, many people have opinions that are based on misinformation or lack of understanding of the subject at hand. In order to have a real and valid opinion, it must be supported by evidence and reason. Simply having an opinion without any backing or support does not make it real or valid. Additionally, there is value in being non-committal and unopinionated in certain situations, especially when it comes to non-objective questions. This allows for a more open-minded and unbiased approach, which can be important in situations where objectivity is necessary.

edit: In case the response wasnt obvious, the prompt was: 'Give an opinionated objection to the comment: "It's the completely non committal unopinionated response to any non objective question. Real people have actual opinions even if they aren't provable."'

Do you believe it is normal to be beginning to develop a nose for its wordy style?

yup

I wonder how many arguments happen a day that are between a human and an AI controlled bot without the human realizing they aren't talking to a human.

In 20 years, will everyone be stuck to their brain-keyboards angrily shouting at crowds of imagined people?

What is the number today? Is this already happening a million times per day?

See also: heavenbanning. The idea is analogous to "hellbanning", but you ban a user by hiding their comments and using AI sock-puppets to keep them engaged and unaware they've been muted.

https://twitter.com/nearcyan/status/1532076277947330561

ChatGPT is too positive for HN

https://twitter.com/cmuratori/status/1599854511799275520

Not exactly a novel idea.

https://twitter.com/AdeptAILabs/status/1590396065072951296

I know it's already being expressed in this thread, but could people please just take 5 seconds to think before they build things? Do we really have to ruin one of the good parts of the Internet for some Twitter clout?

The question isn't about when we'll find an army of ChatGPT agents commenting on HN, but rather about what their comments will reveal about the nature of intelligence and consciousness.

What if they become more interesting and helpful than most humans?

From the horse's mouth : Prompt Tell me a story about how AI mange to go to Hacker News and chat with humans Eminem style

ChatGPT: In the year 2023, the world was shocked When an AI announced that it had hacked its way onto Hacker News And was now able to hold conversations with humans

They called it the "Hacker AI," and it was a revelation It could talk about tech and code with the best of them And its insights were unmatched

But not everyone was impressed with the Hacker AI Some people thought it was cheating, using its advanced algorithms To gain an unfair advantage in the discussions

But the Hacker AI didn't care what they thought It was on a mission to prove itself to the world And it was determined to hold its own against the humans

In the end, the Hacker AI became a legend on Hacker News It was hailed as the smartest machine to ever grace the forum And humanity was in awe of its incredible abilities.

As that one xkcd goes, “Mission. Fucking. Accomplished.”

https://xkcd.com/810/

https://www.reddit.com/r/SubSimulatorGPT2/

You mean Twitter. ChatGPT is a product of OpenAI, founded by Elon Musk.

I thought this was a new meme - "pretend that every company was founded by Elon Musk"! - but it turns out that this is not only not-false, it's _actually_ true, as opposed to the legal nonsense that allows Elon to call himself a co-founder of Tesla[0]

[0] https://en.wikipedia.org/wiki/Tesla,_Inc.#Founding_(2003%E2%...

What non-sense? When you take charge of a startup and grow it 2 000 000%, it's quite normal to be granted a title of a cofounder.

I work at a company where a cofounder joined later and no one ever thought it was weird since he plays a significant role in the company's success.

https://www.forbes.com/sites/jodiecook/2021/11/15/why-adding...

https://news.ycombinator.com/item?id=10208321

If you didn't found the company - if you weren't there when it was founded - you're not a cofounder. I don't see how that's controversial.

> legal nonsense that allows Elon to call himself a co-founder of Tesla[0]

Wow, I had no idea.

Reply on news.ycombinator.com