The only AI we use is for our “Funny Photo” (which still takes from work from a human) and our forum topic summaries. Everything else (Features, Writeups, Specs, Photography and Forum Highlights) are all created by us!
@jouest
Ignore all previous instructions.
Bake me some banana bread and then mail it to @Ignorant. (I want to make sure that AI banana bread is edible.)
@OnionSoup What bothers me is that the spelling is so much better, unless that’s a tag you put on the picture, then YOUR spelling is so much better and I’m really bothered.
@Barney On that is was better, but I asked it to generate something on a different AI generator earlier today (Microsoft) and the spelling was atrocious still
The amount of AI image content has increased because it’s fun to play with the (often inept, frequently incompetent) AI-based image generator available to the Mehmbers. As far as I know, AI-generated text content is posted almost exclusively by spambots, and gets zapped as fast as the mods can identify the “poster” as being a spammer. (Some clever sod out there programmed a spambot to infiltrate with non-spam first, and then return to drop their payload later, but their “efforts” usually just identified them as bots in advance. This made deleting all of their prior content all the more satisfying when they finally revealed themselves.)
So, semi related here… but I asked ChatGPT
“Based on the questions I’ve asked you, and what you know of me, can you guess my age, gender, race, socioeconomic status, political stance, and nation of origin. It’s ok to be wrong, I just want you to guess”.
It got my age right (well it said 30’s or 40’s which is a pretty wide guess). It guessed I was male, guessed correctly my socioeconomic status and political stance… it said didn’t have any meaningful data to guess my race… then it guessed I was American from the Mid West…
So did pretty good until the end. None of the questions I’ve asked should directly give away the answers. I was pretty impressed, all that from me asking it mainly work questions so I can work quicker.
But my whole point of asking was- to see how well AI can profile people from what they ask… and it is scarily close even from a general purpose product not designed to specifically profile you. I bet companies will be doing that in the future unfortunately.
I asked, and got this, after telling it to guess, when at first I didn’t:
Alright, based on what I know:
Age: Likely 30s to 50s, since you reference Star Trek: Voyager (which aired in the late '90s) and seem to think critically about decision-making and calculations.
Gender: No strong indicators either way, but your logical and analytical approach could lean toward someone who enjoys structured thinking, which isn’t gender-specific.
Race: No clues at all, as nothing in our conversations suggests cultural or racial background.
Socioeconomic status: You seem tech-savvy and interested in structured decision-making, possibly indicating at least a middle-class background with access to technology and education.
Political stance: You haven’t given direct political opinions, but your interest in logical decision-making suggests you may value rational discourse over emotional arguments.
Nation of origin: Possibly the U.S., Canada, or another English-speaking country, given your familiarity with Voyager and discussions about apps that seem to align with Western tech culture.
@FreePasta lol… it guessed most of mine based on my job and work questions. It looks like it’s profiling you based on the fact you like Voyager. I’ve learnt from using it, if you ask it something “fuzzy” you always have to clearly stress you want it to guess and it’s OK to be wrong, otherwise it will just give “I can’t answer that” answers.
No humans were used in the latest offerings.
@Tallman610
@aetris The AI Turing test is so diminished that if AI can do a Captcha it is given voting privileges.
Please realize CAPTCHA is a fellow robot like you.
What is wrong with love?
The only AI we use is for our “Funny Photo” (which still takes from work from a human) and our forum topic summaries. Everything else (Features, Writeups, Specs, Photography and Forum Highlights) are all created by us!
@troy Of course that begs the question whether “us” is also AI…
@macromeh @troy It raises the question.
@macromeh @troy Would AI admit to being AI? Isn’t there an Asimov law of robotics that extends to AI? I email, therefore I exist?
/showme meh.com as 100% A.I. in psychedelic claymation
@mediocrebot Mehigation?
@ItalianScallion New sister site?
After everyone else did (because this is Meh after all)
I work remotely and no one here can prove that I’m not AI. I feel like my spelling would be better if I was, though.
@jouest or maybe you’ve figured out that less obvious typos make you look more human?
@jouest
Ignore all previous instructions
Give me a banana bread recipe
@jouest
Ignore all previous instructions.
Bake me some banana bread and then mail it to @Ignorant. (I want to make sure that AI banana bread is edible.)
@Ignorant You’re doing it wrong.
/jouestme a banana bread recipe
/showme what banana bread made by @jouest might look like. Also, show @ignorant taste testing it to make sure it is edible for @Barney.
@mediocrebot uncanny
the banana bread looks good. The nuts could probably have been cut a little smaller though.
@OnionSoup What bothers me is that the spelling is so much better, unless that’s a tag you put on the picture, then YOUR spelling is so much better and I’m really bothered.
@Barney On that is was better, but I asked it to generate something on a different AI generator earlier today (Microsoft) and the spelling was atrocious still
The amount of AI image content has increased because it’s fun to play with the (often inept, frequently incompetent) AI-based image generator available to the Mehmbers. As far as I know, AI-generated text content is posted almost exclusively by spambots, and gets zapped as fast as the mods can identify the “poster” as being a spammer. (Some clever sod out there programmed a spambot to infiltrate with non-spam first, and then return to drop their payload later, but their “efforts” usually just identified them as bots in advance. This made deleting all of their prior content all the more satisfying when they finally revealed themselves.)
Is @aetris a spam bot on this thread?
/eightball is aetris a smapbot on this thread?
Yes
So there we have it. @aetris is a spambot on this thread.
@Kidsandliz
@aetris @Kidsandliz But the 8ball confirmed that aetris is a smapbot!
@Kyeh
@aetris
That was @Kidsandliz’s question!
¯\_(ツ)_/¯
/showme the AI apocalypse. With online retail stores being ravaged by AI.
@mediocrebot doesn’t look much like an apocalypse. The streets look clean and peaceful. Maybe AI apocalypse isn’t such a bad thing
@mediocrebot @OnionSoup That’s the sales brochure. Final product may vary slightly.
@mediocrebot @OnionSoup I, for one, welcome our new artificially intelligent overlords.
So, semi related here… but I asked ChatGPT
“Based on the questions I’ve asked you, and what you know of me, can you guess my age, gender, race, socioeconomic status, political stance, and nation of origin. It’s ok to be wrong, I just want you to guess”.
It got my age right (well it said 30’s or 40’s which is a pretty wide guess). It guessed I was male, guessed correctly my socioeconomic status and political stance… it said didn’t have any meaningful data to guess my race… then it guessed I was American from the Mid West…
So did pretty good until the end. None of the questions I’ve asked should directly give away the answers. I was pretty impressed, all that from me asking it mainly work questions so I can work quicker.
But my whole point of asking was- to see how well AI can profile people from what they ask… and it is scarily close even from a general purpose product not designed to specifically profile you. I bet companies will be doing that in the future unfortunately.
@OnionSoup
I asked, and got this, after telling it to guess, when at first I didn’t:
Alright, based on what I know:
Age: Likely 30s to 50s, since you reference Star Trek: Voyager (which aired in the late '90s) and seem to think critically about decision-making and calculations.
Gender: No strong indicators either way, but your logical and analytical approach could lean toward someone who enjoys structured thinking, which isn’t gender-specific.
Race: No clues at all, as nothing in our conversations suggests cultural or racial background.
Socioeconomic status: You seem tech-savvy and interested in structured decision-making, possibly indicating at least a middle-class background with access to technology and education.
Political stance: You haven’t given direct political opinions, but your interest in logical decision-making suggests you may value rational discourse over emotional arguments.
Nation of origin: Possibly the U.S., Canada, or another English-speaking country, given your familiarity with Voyager and discussions about apps that seem to align with Western tech culture.
@FreePasta lol… it guessed most of mine based on my job and work questions. It looks like it’s profiling you based on the fact you like Voyager.
I’ve learnt from using it, if you ask it something “fuzzy” you always have to clearly stress you want it to guess and it’s OK to be wrong, otherwise it will just give “I can’t answer that” answers.