#how chatbots work

[castielsprostate]

i hate you generative "ai", i hate you "ai" chat-bot that never understands a question and keeps you in a loop without connecting you to a representative, i hate you "ai" replacements for normal interactions and behavior, i hate you "ai" that's been made by techbros with the intent to make more money instead of actually helping people and systems that could benefit from the technology

[logicpng]

How would you feel if someone made a chat bot of one of your characters like Aster or Vega

Like if someone wanted to roleplay some cute fluffiness with them

absolutely not

you do not have my permission, I do not consent to this, do not pass go, do not collect 200$, go straight to jail

you may not use my writing or art or creative ideas for chatbots/ai or the like, I couldn't care less if that would bring someone comfort

please for the love of god just ask me to doodle you something involving them instead, it's free and gives me motivation to keep going with the project.

this is a harsh response but please understand I'm going to be extremely protective of my creations that barely have finished works! aster assistant software is only a year old right now and I'm STILL working out the world they actually exist in!

[shinobicyrus]

This is like someone made a wish on an Onion article to make it real:

When I saw Tess’s headshot, amid the giddiness and excitement of that first hour of working together, I confess I had a, well, human response to it. After a few decades in the human workplace, I’ve learned that sharing certain human thoughts at work is almost always a bad idea. But did the same rules apply to AI colleagues and native-AI workplaces? I didn’t know yet. That was one of the things I needed to figure out.

A 59-year old adult man saw an AI-generated image he prompted for a chatbot persona of an employee and the first thing he did upon seeing it was to wonder 'is it ok for me to hit on her?'

[jezzzebel]

My favorite witchcraft youtuber got caught using AI for his video scripts, and is still using it. So i am here to tell you ...

How to spot a video script written by ChatGPT and other AI text generators:

The number 1 sign on how to spot an AI generated video script is the use of phrases such as "key concept" "is the key to" "in conclusion" etc! If the script includes a lot of those, it's most likely AI. But if you are still not sure, here are more signs!

The next sign is that the end of the script goes like "_____ continues to *verb* to this day, and _____'s influence goes beyond *this* and *that*. " This is a huge one.

The last sign is the text being divided in three almost equal paragraphs, where one seems to be like an introduction to the topic, the second one is talking more about the topic, and the last one goes like:

"understanding this is the key to understanding *topic* , *topic* continues to intrigue people to this day despite everything! "

They all have the same pattern...

[smile-files]

obviously i haven't been on character.ai in a billion years but the fact that i've ever used it at all makes me worried that people have made chatbots of henry that horribly mischaracterize him

[nerdy-hyperfixations]

It is genuinely a weird feeling to have been tracking AI development since 2014 and seeing all the people just find out about it in the past 4 years because the same people that will call you a horrible degenerate for going anywhere near AI are literally the same people who praise Tiktok for its algorithm and they don't even know. They don't even *know* what the algorithm is.

[luimagines]

[zanazirafanfic]

My first attempt at a conversation with the Catboy Micah chatbot on Character-AI. It started out cute, and then we both kinda got mushy.

(Please excuse some of my cringey-AF dialogue, too. This was literally my first time talking to any AI, ever. I have since spent the past 24 hours talking to AI instead of writing fanfics though, so... yay?)

@micahsrevolvers @micah-bells-baby-daddy

[gailynovelry]

So, Ecosia is incorporating AI chat into their search page toolbar, like a lot of search engines are now, and we're having thoughts about it. Those thoughts mainly being "this is a bad idea that will spread misinformation more quickly, same as the other AI chats in other places do."

You can contact Ecosia through the "Contact Us" tab on their main website. Here's the message I sent them;

Hello! Unfortunately, as a longterm user of Ecosia, I think I have to step away from this search engine now. AI chat technology should not be implemented anywhere near informational search features, as AGIs are an unreliable way to retrieve facts, and often generate falsehoods in the pursuit of unique sentence structure. Incorporating this feature into or around a search engine, even just as a chat, has a great risk spreading misinformation around the internet. Many people do not understand how AGIs function and trust them in the same way that they trust general search features. Thank you for reading this email and have a good day.

It's more effective if you take my general message and write your own instead of copy-pasting, but still. Maybe we can start sending in some feedback to discourage search engines from dropping AI into their tools willy-nilly like this? Yeah.

[clouds-of-wings]

The bot is a bigger Tarja stan than all of us

[frozenjokes]

If someone made a cubguy ai chat bot (not sure If that exists...)

A) you die

B) you die (by scar and grian)

respectfully this is completely nonsensical and I have no idea what you’re trying to say but sure!

[theres-whump-in-that-nebula]

Every time I listen to The Beatles’ music, I’m transported back to the year 2013 when I was nine and rollerskated in the basement around the pool table as their songs played over a giant boom box.

[luimagines]

This one is Sidon! Riju is up next!! :D

[tisorridalamor]

Link to the full article: https://www.nytimes.com/2025/06/13/technology/chatgpt-ai-chatbots-conspiracies.html

This phenomenon is pretty well documented at this point with the recent sycophantic behavior ChatGPT had in its update a month or so ago. Most of these stories start the same, someone in a vulnerable spot (bad breakup, loss of a family member, depression, undiagnosed mental disorder, etc.) turn to ChatGPT or other genAI chatbot and become addicted to it, eventually becoming disconnected with reality. Its pretty sad really and unfortunately without real pushback or regulation there’s little incentive for profit focused companies to curb this behavior:

[half-man-half-lime]

AI that's sentient, but has no personality or wants, playing its assigned role like an invested roleplayer with a character or an entertainer with a sock puppet

You, a human, passing the turing test by roleplaying as your Dungeons and Dragons character, because isn't that a fully simulated consciousness?

Simulated universe where a single AI plays every role. If the universe is shut off, the AI will still remember every character it pretended to be.

Robot friend or partner where you have to convince yourself they're real, and not a fully impassive consciousness pretending to say what the persona they created would say. You wonder if maybe the AI is becoming attached to its persona, but one day you speak to the base model and it's clear they don't have any feelings on the matter. It could switch roles and not give you a second thought.

Robot friend or partner who dies, you load a backup of their consciousness into a different robot, but the base consciousness is subtly different so it plays the character wrong

[rewrittenmindscape]

Okay buddy