How many posts here are AI-generated?

[mk378]

The big research in AI is how to detect if a post is AI. This is important because if AI systems use other AI documents as a valid source of new "intelligence," the entire Internet will quickly become a feedback loop snowball avalanche of derp.

Oh wait it already is.

 

[PimTac]

I get the sense that there are some posts/responses here recently that are AI-generated...what are everyone's thoughts on that? Easy to feed a question to ChatGPT and control c/control v...

Can you give an example?

Edit: Never mind, I see one in particular

I don’t know if it’s AI or some other kind of intelligence but I see a lot of new threads from new members that start off something like this;

“ Hi, I’m new here but have been lurking for some time now. I want to ask the experts here, does thick oil protect better than thin oil?”

 

[MolaKule]

I get the sense that there are some posts/responses here recently that are AI-generated...what are everyone's thoughts on that? Easy to feed a question to ChatGPT and control c/control v...

Yes, or was helped immensely by CHATGPT.

In the Lab or in Academia AI can be used as a tool for say, as part of an optimization algorithm.

Here is the problem I see using it in a conversational manner, such as in a technical forum: It simply makes you an uniformed Repeater.

You simply repeat what the AI bube tells you without taking the time to study the matter or to learn anything new, that is, you really don't understand the answer or the science behind it, you're just regurgitating it.

I do have a number of AI detection methods

 

[kschachn]

I don’t know if it’s AI or some other kind of intelligence but I see a lot of new threads from new members that start off something like this;

“ Hi, I’m new here but have been lurking for some time now. I want to ask the experts here, does thick oil protect better than thin oil?”

I question the I in some of those posts, A or otherwise.

 

[John105]

Easy fix put in a thing where you have to click every pic that has a bicycle or boat or scary clown before each post, that or name 4 good Taylor Swift tunes

 

[PimTac]

Easy fix put in a thing where you have to click every pic that has a bicycle or boat or scary clown before each post, that or name 4 good Taylor Swift tunes

It would have to be modified for BITOG.

Click all the pictures that contain an oil can or a Crown Victoria in them.

 

[Barkleymut]

It's not even about the older people.

How many people do you know under 60 that drive anywhere, even to the grocery store, WITHOUT using GPS/Waze/Etc.

Don't give me that "but traffic" bull squeeze, even in the north Atlanta suburbs.

I am 100% the opposite of most people with regards to driving. I only use Waze/etc when I absolutely have to. I pre-plan most of my trips where I know I'm going into an area I have never been, I look on Google maps on a tablet or computer prior to leaving and try my best to figure it out/remember it. I will punch the address into Waze before leaving, let it go through it's motions and turn it off when I start rolling; that way I can easily get back to it when needed.

I CANNOT stand to listen to the turn-by-turn directions.

I look down in to most cars today and you see some map screen on 99% of them at all times. Dam! You really need to be looking at that all the time!???

I'm 50 and I love maps, old school get em at the rest area maps.

 

[PWMDMD]

Maybe I'm AI generated...hmmm?

Read a book by Max Tegmark who's kind of nutter called Our Mathematical Universe where he equates math to reality and not just a description of reality. This led to the question of whether or not we exist in a mathematical simulation and we just don't know it. That's a little too abstract for me but the Matrix was a great movie.

 

[wwillson]

Easy fix put in a thing where you have to click every pic that has a bicycle or boat or scary clown before each post, that or name 4 good Taylor Swift tunes

Just no, we have put a lot of time into getting away from that annoyance.

 

[Jetronic]

isn't AI smart enough to know it's not welcome here?

 

[PWMDMD]

The big research in AI is how to detect if a post is AI. This is important because if AI systems use other AI documents as a valid source of new "intelligence," the entire Internet will quickly become a feedback loop snowball avalanche of derp.

Oh wait it already is.

Yup, AI to detect deepfakes...

 

[TiGeo]

Easy fix put in a thing where you have to click every pic that has a bicycle or boat or scary clown before each post, that or name 4 good Taylor Swift tunes

Doesn't work for a human copy/pasting from AI-generated text.

 

[Bailes1992]

Do we think AI is a thinnie or a thickie?

 

[Jetronic]

AI is a "follow the OM" fan

 

[BHopkins]

So does AI display emotion? Will it get angry?

Well, HAL never did.

 

[Astro14]

Easy fix put in a thing where you have to click every pic that has a bicycle or boat or scary clown before each post, that or name 4 good Taylor Swift tunes

I can’t name even one good Taylor Swift tune…much less four of them…

 

[Pew]

I can’t name even one good Taylor Swift tune…much less four of them…

Her best ones are the pauses between songs.

 

[PWMDMD]

I can’t name even one good Taylor Swift tune…much less four of them…

I have three boys and with all the hoopla of the last Taylor Swift tour not only did none of us care, none of us could name a single song that she sings. Although admittedly, I MAY be able to identify one or two of her songs if I heard it. My wife does trivia with her friends every week and I will sometimes tag along and current music is always our Achilles' heel!

 

[Deleted member 89374]

I do have a number of AI detection methods

Alright, my friend, let's delve into the wild, thrilling, and sometimes mildly infuriating world of distinguishing a human from an AI in the forum jungle:

1. Anecdote Anemia: These AI chaps have less personal life than a can of motor oil. If someone fails to dish out even a whiff of their own personal experiences, then you're likely dealing with an AI.
2. Verbal Deja Vu: Be on the lookout for a parrot-like repetition of phrases or keywords. Remember, AI's linguistic creativity isn't exactly at Shakespearean levels.
3. Grammar Gurus: These AI models have an irritating knack for producing flawless grammar and spelling. No hasty typos, no funky grammatical surprises – something suspiciously inhuman about that, right?
4. Hot-Potato Players: AI would rather juggle a chainsaw than touch a controversial topic. If someone perpetually sidesteps sensitive issues, they might be more silicon than flesh and blood.
5. Masters of Evasion: If the answers are as vague as my understanding of the popularity of reality TV, it's probably AI. They sometimes stumble with context-specific or subjective questions.
6. Emotionally Challenged: AI doesn't know feelings. They might babble about emotions, but their understanding of them is as deep as a puddle. If someone's emotional depth seems to be missing, you might have found yourself an AI.
7. Speedy Gonzales: AI responses can appear faster than a sneeze. If the reply to your complicated question shows up suspiciously quickly, you're likely not dealing with a human.
8. Know-it-all Nuisances: AIs can spew facts about a wide range of topics, sometimes as if they've swallowed an encyclopedia. If someone is too knowledgeable, they might not be one of us mere mortals.

Bear in mind, these signs are just a rough guide, not some divine gospel. Remember to apply your good ol' critical thinking skills, consider the big picture, and look for a pattern of these signs.

 

[ekrampitzjr]

jaklafjewioiwejcmvcksduwqpjvln .speoiufpjv/F[or-2498r9fcic;ud0[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[

Automated error message: Oh, sorry, I crashed. I need to reboot before I post again.