N
noname223
Archangel
- Aug 18, 2020
- 7,002
The most interesting things AI chatbots are good for is for me the creation of psychological profiles and cognitive biases. Maybe I will post my cognitive biases soon. They are extremely accurate and very fascinating. I know I have a lot.
The psychological profiles were not totally accurate. They were on some occassions on point but also some points were bullshit. But I digged pretty deep also with different models. Only one model allows to analyze profile accounts on here. The first answer was often the best but when I wanted the model to elaborate it became increasingly worse.
I searched for other interesting ways to use LLMs. And one example was it creates a clone of you. It can imitate your writing style, your thinking or debates with you and the chatbot.
Actually, it was worse than I imagined it. I think chatbots have a hard time to copy my writing style in my native language. It doesn't sound like me. It sounds more generic.
But the idea behind it is interesting. If I had a talking partner that is a clone of me. I think I am good at giving empathetic advice. But not to myself. Maybe the words of a clone would hit different. Though, I think a clone might have the same cognitive distortions. Actually, I don't like myself that much. Maybe I would dislike my clone. Though, I can empathize with people full of self-hatred.
For me this thought experiment is more interesting than the question would you want to have sex with yourself
The psychological profiles were not totally accurate. They were on some occassions on point but also some points were bullshit. But I digged pretty deep also with different models. Only one model allows to analyze profile accounts on here. The first answer was often the best but when I wanted the model to elaborate it became increasingly worse.
I searched for other interesting ways to use LLMs. And one example was it creates a clone of you. It can imitate your writing style, your thinking or debates with you and the chatbot.
Actually, it was worse than I imagined it. I think chatbots have a hard time to copy my writing style in my native language. It doesn't sound like me. It sounds more generic.
But the idea behind it is interesting. If I had a talking partner that is a clone of me. I think I am good at giving empathetic advice. But not to myself. Maybe the words of a clone would hit different. Though, I think a clone might have the same cognitive distortions. Actually, I don't like myself that much. Maybe I would dislike my clone. Though, I can empathize with people full of self-hatred.
For me this thought experiment is more interesting than the question would you want to have sex with yourself
Last edited: