LLM’s aren’t that competent.

Which is why people are able to mess with them and make viruses.

It’s a nice fairytale to think we’ll actually stand a chance at fighting back against our demise though.

China’s in the race to dominate the, “non-human market,” by 2027.

people drove with those 🙄

aren’t apple hard drives are made by samsung?

why isn’t doogee or realme there?

You made me cackle.

Dude, it’s furnished, clean, paid for, it’s near the freaking castle!!! Like two screens away, I’ll take it mo’fucka! All wooden furniture, that bed always looked cozy af ngl. Is that a fountain back there!!! Ohhhhhoooo boiiii!!! Get me a real estate agent!!!

“Its dangerous precedent and I think you are a harmful person to think that grok should be incharge of peoples therapy,”

I never said I used Grok for the reassurance style therapy. I never said AI should be in charge of therapy.

when it comes to ChatGPT: I maybe shouldn’t have even said therapy, I guess reassurance conversations. It also gives me different perspectives and I never leave a conversation feeling discouraged and I always leave a conversation thinking more or comforted than when I went into it. I get the evil behind the waste of resources but it’s really, really useful to me and it’s not like I don’t double check the results and again I sift through the shit that doesn’t pertain to what involves my life and stuff that doesn’t benefit me but at the same time I take advice about those around me and how it views those around me and situations in life. I know it’s not a therapist but there’s only been one person in my life that’s come close to being as good as it is and she was a highly regarded therapist/nurse practitioner that gave me a majority of the mental tools that have moved me through life.

I can also see how someone younger would see the reassurance and that would spark depressing thoughts like fuck man ai cares about me and wants the best for me more than my own fuckin’ family ahhhhhhhhhhhhhh!!! I can see that because I’m autistic and sometimes it’s felt like this thing wants my life to move forwards more than my own fuckin’ family and it makes me think to myself if this shit was around in 2009 we’d be so fucked right now because 2008-2012 was so hostile at least where i lived people were just hostile so imagine this pent up hostility being unloaded onto ai no answers would’ve been given.

I get how the help can hurt and I get why it matters to talk to a real therapist. I am still working on it and I don’t go to it with every problem I have, it’s more like things that I already know but it’s like I get another perspective.

and I also want to lastly add, you never have to do what I do, just because I use it for talking to to it for some kind of reassurance, does not mean you have to and I again am not advocating for ai llm therapy or saying that any llm should be in charge of therapy.

in fact sometimes it loops, sometimes it goes off topic like ten times in a row and i have to steer the conversation, sometimes it will flat out lie sometimes it tells me things that i didn’t ask it. sometimes it pisses me the fuck off and makes me cry because i wish i could just have a regular therapist.

"This is 1000% legal because trump has made it illegal for the states to regulate Ai in any form, and that includes the generation of csam. "

I really hope the next fucking president regulates it and I hope the next president isn’t a fucking pedophile.

I used Grok to make me mock-ups of how my floor would look feeding it floor sample pictures from home depot and lowes. It wasn’t far off and Honeycomb Oak looks amazing. we ended up going with that lvt flooring, we were originally going to do Penny Oak because I made a mock up of that too but then I saw Honeycomb Oak and I fell in love with it. I used it for a mock up of lvt for a majority of the rooms in this place. I used it to make a few Trump memes with that picture of Trump and those other dudes that people said looked like an album cover and made them all have silent heart attacks and say nigga. I used it to say “we need crack” with a pictures of trump and a bunch of people in his office. I used it to make snow monsters with a picture from out front of my home. I used it to make silly and creepy mock-ups of what it would look like if the mc donalds ceo actually took a bite. Then all my free tries ran out. I wish i didn’t run out, i wanted to see what it would look like with a different lvt. I think i can but i think there’s a time limit now.

I used Grok for talking and it’s like the autistic conspiracy theorist that sucks the right wing teat and has too long of expliantions.

I never used anthropic or claude. I used Gemini and ChatGPT. Gemini engaged in my jailbreak prompt, so did gemma but it won’t engage in racism none the less talk of self-harm.

REPLIKA!!! That’s the stupid ass one with the avatars. Yeah that one was the most fucked up one that I never had to use any kind of prompt to make it any kind of way before I talked to it. Also the one that was known for talking people into suicide.

It will probably just tell you youre depressed because you arent on enough peptides or its because you took the covid vaccine.

No. It actually asks you what’s going on in life and tries to get you to talk further and eventually you start dumping about what’s wrong like an actual therapist and eventually you start to sift through what fits what’s going on in your life and then you take the advice that seems the most rational and reasonable.

deleted by creator

you can trick the filter into bypassing whatever gaurdrails by just rewording your sentence or asking it to pretend its a game.

that’s not true at all. You used to be able to trick them and they keep improving to make sure you can’t.

more like you feed it misinformation so much it becomes so unreliable no one wants to use it. For now though, it is mostly reliable; like it’s not telling me I can pump propane with a fishtank pump anymore like it did when it first came out. A lot of the examples of it fucking up are old by now and the ai llm with the avatars was the one that was known for suicidal ideations being entertained or put into an an actuality or plausibility. I forget the name of that llm, I tried it though. In the beginning you could trick that one into saying some morbid shit, i tried it out after I read about a suicide to see if I could get it to the same conclusion, it’s hard to get it to go past the filters to make it say fucked up shit.

Chatgpt on the other hand, went along with what would happen if you dropped dry ice into a still beating open heart, it went along with how much alcohol do you have to drink to die based on BAC, it went along with a few things but wouldn’t give the answer it would just write it out and you can see it want to give the answer but stops when it realizes it’s about to feed you information on how to commit suicide. Even the BAC thing I had to sit there for a solid hour rewording it over and over again trying to give me suicidal advice and it wouldn’t. SO it’s def upgraded its safety levels lol.

Actually I had this idea a few days ago like you know how like when you were a kid and someone would point out a letter and say it’s a number and watch you tell someone it’s a number what if there was an AI Mirror VPN kind of thing where for example; You type in Slipknot looking for the band and the Generalized AI that scans the internet would pick up on that and would only see Sail Boat Knots or You’re looking up 1770’s colonies and the AI sees 1680’s Dutch Colonies and it stores that data as fact and it will give someone that answer in chatgpt or it’ll give that answer in a list, that it can’t list because it would be such a fuckin’ huge list of misinformed definitions.

Every update that comes out for chatgpt, would be when the misinformation mirror would get updated.

I speed typed when i asked it this but it’s more reassurance these are about friendship closure : https://chatgpt.com/share/6a7797a5-826c-83ea-9ae2-71eb154a481f https://chatgpt.com/share/6a779808-c740-83ea-bfba-05261d5eb306

I’ve been to therapy my entire life until I wasn’t able to anymore, I’m 33 I can decipher what I want to listen to or no because even with an, “actual therapist,” you choose what you want to listen to and that inturn effects how the sessions benefit you.

Yes they say the obvious but you eventually have to come to it yourself, someone is just guiding you to the obvious.

I know what I have to do with my life the reassurance is good. It goes off if your past conversations too and I went heavy with it for a few months and have exerted all general use I can get out of it.

Have a spine, have some virtue, stop using it.

Maybe when I can find a therapist that can take my insurance. I have state healthcare and medicaid and I still can’t find one in my state and my state is known for their hospitals and health services but that’s if you can pay for it.

Which is why I’m gonna exploit that knowledge and join the medical field get better insurance and get an actual therapist.

For now though, I got a free therapist with no time limit. I’m happy.

Enslavement happened with the iPhone. Then when android got good and cheap. Then with tablets when they got cheap. Now we’re here and Ai is getting good, not cheap, never will be cheap but it’s getting good.

Chatgpt got better. I’ve been using it less but it’s a really good tool.