As much as I like some good bias confirmation, the responses of the models is more of a reflection of the data they were trained on than a logical conclusion they came to. One could just as easily train a model to make the opposite opinion given the apropriate(incorrect imo) data. Which I guess may be telling about the in depth discussion of bitcoin vs fiat on the internet, that was scraped to train them. On avarage pro fiat people may be more dismissive and not as willing as bitcoin/stable coin users to discuss the topic in detail.
As much as I like some good bias confirmation, the responses of the models is more of a reflection of the data they were trained on than a logical conclusion they came to. One could just as easily train a model to make the opposite opinion given the apropriate(incorrect imo) data. Which I guess may be telling about the in depth discussion of bitcoin vs fiat on the internet, that was scraped to train them. On avarage pro fiat people may be more dismissive and not as willing as bitcoin/stable coin users to discuss the topic in detail.