Raw LLM Responses

I’m frankly much more concerned about the use of AI for spying and mass surveillance, as well as military operations and propaganda, much more than cultural bias. If AI development is made open-source, there’s no reason that cultures in non-Western countries can’t make AI that are more aware of their cultural norms, while still having access to the same algorithms as other AI that make it function so well for information and language purposes. The problem is, when a bunch of technocrats and corporations own it, the biases are likely to devalue and even insult diverse opinions, tending to protect the interests of big corporations. Lowering prescription drug prices, lowering the incidence of over-prescription of unnecessary or even inappropriate medicines or devices (we’ve all seen the recall and class action lawsuit commercials that result from pushing something too much and too soon), and consuming less junk instead of constantly spending might not be treated as equal solutions to big problems, as corporate-friendly and government-friendly solutions. I’m a lot more concerned that AI will be used as automatic advertisement basically, and automatic surveillance. Cultural preferences and history are easy enough information to expose an AI to— training an AI not to be a simp for the corporations and governments that fund their research is gonna be a much bigger task.