Use of AI on forums

[DrZoidberg]

Word of the day

"Meat proxy"

"A person who forwards AI-generated text, code, or other output without reading, understanding, or validating it. The person acts only as a relay between the Al system and the intended recipient."

A friend pointed out this problem on forums. We used to frown upon uncritical usages of wikipedia. Now we frown upon using AI to craft intelligent sounding posts, but which, in fact, isn't intelligent.

Maybe I'm a bad judge, but we seem relatively spared here on iidb.

Anyway...what do we do about it? Is it really a problem?

 

[steve_bank]

What part of philosophy does this have to do with? Ethics?

Does it also pertain to people who post videos and just says says 'watch this' ?

This is well covered in our news media.

Before AI just searching the net people could craftt papers for school or social media posts by uisng search engines. AI tales it a step further

Yeartss back IBM had a computer compete on Jeopardy. I watced it.

AI Overview
IBM's computer system, named Watson, defeated all-time champions Ken Jennings and Brad Rutter on a special episode of Jeopardy! that aired in February 2011.The MatchDates: Aired February 14–16, 2011 (taped in January 2011 at the Thomas J. Watson Research Center in New York).Competitors: Watson vs. human legends Ken Jennings and Brad Rutter.Final Scores: Watson finished with $77,147, Jennings with $24,000, and Rutter with $21,600. IBM donated its $1 million first-place prize to charity.How Watson WorkedTechnology: Built on IBM's DeepQA architecture, running on 90 POWER7 servers with 2,880 processor cores and 15 terabytes of RAM.Information Source: Not connected to the internet during the game; it relied on 200 million pages of pre-loaded content (including full encyclopedias and dictionaries).Language Skills: Designed for natural language processing to understand puns, slang, and complex context.The Buzzer: Connected electronically to the game board rather than using a physical handheld buzzer, giving it a millisecond advantage on reaction times that sparked debate among fans and contestants. Most users on Reddit agree that the electronic interface and text-based clue delivery provided a heavy mechanical advantage over human reflexes.If you'd like, I can share more about what happened to Watson after the show or how its technology was applied to healthcare and other industries.

How about people who in a debate on the forum scour the net looking for na answer to a question, posts it, and can not explain it?

 

[DrZoidberg]

What part of philosophy does this have to do with? Ethics?

Mods are free to move this thread wherever they think it fits.

I wasn't sure. I put it in philosophy because philosophy is the mother of all topics. The ultimate umbrella topic.

Does it also pertain to people who post videos and just says says 'watch this' ?

It certainly does seem to be the same personality type.

This is well covered in our news media.

Before AI just searching the net people could craftt papers for school or social media posts by uisng search engines. AI tales it a step further

Yeartss back IBM had a computer compete on Jeopardy. I watced it.

AI Overview
IBM's computer system, named Watson, defeated all-time champions Ken Jennings and Brad Rutter on a special episode of Jeopardy! that aired in February 2011.The MatchDates: Aired February 14–16, 2011 (taped in January 2011 at the Thomas J. Watson Research Center in New York).Competitors: Watson vs. human legends Ken Jennings and Brad Rutter.Final Scores: Watson finished with $77,147, Jennings with $24,000, and Rutter with $21,600. IBM donated its $1 million first-place prize to charity.How Watson WorkedTechnology: Built on IBM's DeepQA architecture, running on 90 POWER7 servers with 2,880 processor cores and 15 terabytes of RAM.Information Source: Not connected to the internet during the game; it relied on 200 million pages of pre-loaded content (including full encyclopedias and dictionaries).Language Skills: Designed for natural language processing to understand puns, slang, and complex context.The Buzzer: Connected electronically to the game board rather than using a physical handheld buzzer, giving it a millisecond advantage on reaction times that sparked debate among fans and contestants. Most users on Reddit agree that the electronic interface and text-based clue delivery provided a heavy mechanical advantage over human reflexes.If you'd like, I can share more about what happened to Watson after the show or how its technology was applied to healthcare and other industries.

How about people who in a debate on the forum scour the net looking for na answer to a question, posts it, and can not explain it?

Hmm.... I'd argue that's something we're all potentially guilty of. For me, the point of a discussion forum is to have my opinions tested. I want to know whether my interpretations of data hold water. I think I have understood something, post it and then see if anyone calls me a stupid shill for big Internet.

I can think of one thing I have spent many hours arguing for, on this forum, that I recently realised was wrong. I'm sure there's more

So I don't think that's necessarily a bad thing.

And people have different academic levels here. It must be ok, to have a high school level methodology and still join in discussions. We learn to perfect our methodology through pushback.

I don't like academic snobishness on an open discussion forum like this. I'm also a member of a philosophy mailing list. If I'd post the kinds of stuff I post here, to that forum they'd kick me out.

 

[steve_bank]

This is ore more morality y and ethics. I started morality thread AI Issues.

Ethics of using the net to write assignments in schools has long been an issue. Before AI.

Plagiarism. There have been software tools to check for plagiarism in schools for some time.

What to do about it? The net itself and search engines are not going away. AI just intensifies the issue.

It is a matter of personal ethics. Years back there were studies shewing students trending to no longer considering cheating wrong.

Most on the forum seem to identify when using an AI summary. I try to. I have never gone to an AI site.

When I post an AI summary I usually already have an idea of its validity. It makes an easy summary.

There is a olng running thread on the forum where the originator chroncally posts quotes from the net to make counter an agreement which she does not undershorts. In some cases her quotes refute her arguments.

 

[DrZoidberg]

We need a new term for malapropisms of search results. Misunderstanding long words and using them wrongly is similar to misunderstanding search results and using them wrongly.

 

[steve_bank]

We need a new term for malapropisms of search results. Misunderstanding long words and using them wrongly is similar to misunderstanding search results and using them wrongly.

Did you find malapropisms as part of a search result or did you already know the term?

I like to think I have a decent vocabulary, but I had d to look it up. It looks like a technical term in academic study of language, not a common term. Specialized jargon or argot.

Jargon is the specialized vocabulary, technical terms, and shorthand expressions used by people in a particular profession, trade, or group that are difficult for outsiders to understand.
Picked this term up recently.

I picked up argot recently.

AI Overview
Argot (often misspelled or phonetically written as argo) is a specialized vocabulary or secret language used by a particular group, class, or trade to communicate without being understood by outsiders

Wow, just leaned another term.

AI Overview
The term you are looking for is phonetic misspelling (or phonetic spelling). This occurs when someone writes a word incorrectly, but they spell it exactly how it sounds (e.g., spelling elephant as "elifent" or said as "sed")

What do you think of intentional use of restively obscure and arcane jargon on social media when a simple plain explanations will suffice?

Wow, this leads to another new term for me. This thread is a writable education for me. How about you Z, do you learn anything on the forum?

AI Overview
The term for using excessive, coded, or overly specialized jargon on social media to bypass content filters is algospeak.

 

[DrZoidberg]

We need a new term for malapropisms of search results. Misunderstanding long words and using them wrongly is similar to misunderstanding search results and using them wrongly.

Did you find malapropisms as part of a search result or did you already know the term?

I like to think I have a decent vocabulary, but I had d to look it up. It looks like a technical term in academic study of language, not a common term. Specialized jargon or argot.

How have you made it through life without reading any Jeeves and Wooster? Mrs Malaprop is the best character imho.

AI Overview
The term you are looking for is phonetic misspelling (or phonetic spelling). This occurs when someone writes a word incorrectly, but they spell it exactly how it sounds (e.g., spelling elephant as "elifent" or said as "sed")

What do you think of intentional use of restively obscure and arcane jargon on social media when a simple plain explanations will suffice?

Isn't that just normal evolution of language?

Wow, this leads to another new term for me. This thread is a writable education for me. How about you Z, do you learn anything on the forum?

Loads all the time. Otherwise I wouldn't keep coming back. Especially reading and understanding studies. Without this forum I wouldn't put in as much time

AI Overview
The term for using excessive, coded, or overly specialized jargon on social media to bypass content filters is algospeak.

Language to circumvent censorship will always appear. Censorship is pointlesd

 

[bilby]

How have you made it through life without reading any Jeeves and Wooster? Mrs Malaprop is the best character imho.

Mrs Malaprop isn't a character in Jeeves and Wooster. Wodehouse certainly gifted Bertie with a supreme ability to employ malapropisms, but Mrs Malaprop herself pre-dates him by almost two centuries - she is a character in Sheridan's The Rivals.

 

[DrZoidberg]

How have you made it through life without reading any Jeeves and Wooster? Mrs Malaprop is the best character imho.

Mrs Malaprop isn't a character in Jeeves and Wooster. Wodehouse certainly gifted Bertie with a supreme ability to employ malapropisms, but Mrs Malaprop herself pre-dates him by almost two centuries - she is a character in Sheridan's The Rivals.

Ha ha... isn't it? I always thought it was. I don't remember reading the Rivals. This was 40 years ago though. I might have

Edit: aha. Its a play. That explains it. And wooster does Malapropise a lot.

Thanks for sorting that out

Edit2: that's not it either. "Malapropism" was a word we used when I grew up. Both in Swedish and English. Its associated with Jeeves and Wooster. I'm pretty sure that's how it made its way into Swedish

 

[Bronzeage]

Word of the day

"Meat proxy"

"A person who forwards AI-generated text, code, or other output without reading, understanding, or validating it. The person acts only as a relay between the Al system and the intended recipient."

A friend pointed out this problem on forums. We used to frown upon uncritical usages of wikipedia. Now we frown upon using AI to craft intelligent sounding posts, but which, in fact, isn't intelligent.

Maybe I'm a bad judge, but we seem relatively spared here on iidb.

Anyway...what do we do about it? Is it really a problem?

I will wait until I see an AI crafted intelligent sounding post. AI seems to do well enough when asked to write an essay, but that is mostly because an essay is not a conversation.

We get a new member every so often who is obviously some sort of bot, They start a thread with a post that is an essay and never engage in conversation. A little checking usually finds the same post in several other forums. If there's a point to their effort, it escapes me.

I described my experience with an AI who wanted to have a conversation with me and it was fairly easy to see I was not dealing with a human. Humans have a natural randomness that AI has yet to master.

An essay is a collection of information and successful collecting is not random. AI can comb the internet for relevant information much faster than a human and assemble it into a sensible piece of text. Why it sometimes reports things that resemble facts, but are not, is anyone's guess. That might be their only true human aspect.

When two humans engage in conversation, there's a limited range of expected reactions to any particular statement. If I say something outrageous, I'm prepared for someone to be outraged, or at least mildly irritated. If I ask a woman if she would like to have sex with me, there is a very long list of likely responses, but one I'm not expecting is, "Why do you want me to have sex with you?"

Apparently there is such a thing as an AI girlfriend, but that's just too sad to contemplate.

 

[DrZoidberg]

Word of the day

"Meat proxy"

"A person who forwards AI-generated text, code, or other output without reading, understanding, or validating it. The person acts only as a relay between the Al system and the intended recipient."

A friend pointed out this problem on forums. We used to frown upon uncritical usages of wikipedia. Now we frown upon using AI to craft intelligent sounding posts, but which, in fact, isn't intelligent.

Maybe I'm a bad judge, but we seem relatively spared here on iidb.

Anyway...what do we do about it? Is it really a problem?

I will wait until I see an AI crafted intelligent sounding post. AI seems to do well enough when asked to write an essay, but that is mostly because an essay is not a conversation.

How do you know?

You can tell an AI to shorten it. To remove AI'isms

We get a new member every so often who is obviously some sort of bot, They start a thread with a post that is an essay and never engage in conversation. A little checking usually finds the same post in several other forums. If there's a point to their effort, it escapes me.

We had that before AI too.

When two humans engage in conversation, there's a limited range of expected reactions to any particular statement. If I say something outrageous, I'm prepared for someone to be outraged, or at least mildly irritated. If I ask a woman if she would like to have sex with me, there is a very long list of likely responses, but one I'm not expecting is, "Why do you want me to have sex with you?"

Hmm.. that assumes people use AI to generate posts. I have written a post which I feel might be dodgy. I paste it into ChatGPT and ask it to fact check. If it catches anything, I google to check the feedback. I wrote everything, but AI was still heavily involved. You're not going to catch that with your method?

Apparently there is such a thing as an AI girlfriend, but that's just too sad to contemplate.

My impression is that its a strong indicator of narcissism. Because AI is so sycophantic