“LLMs do not have any notion of what is and isn’t true - Walid Saba”.
So if they are trained to become biased on certain theories-hypotheses, beliefs, customs, decisions, strategies, all different aspects of human thought and action, they may tremendously influence choices we make as humans and navigate us to the wrong direction, or better say the direction that others chose for us to follow. You can achieve this with many other ways such as “buying” conscience of others. Well, we have just discovered another way to do this smarter.
It’s all about free will, about how easily we submit ourselves to automated processes, how easily we trust a system or people and we waive our right and responsibility to take critical decisions for our lives and lives of others. We have already seen how dangerous this is with social media and how young people can be easily manipulated. With LLMs there can be an even more sophisticated deceit at a personal and/or society level.
The issue is not about how to regulate AI and LLMs, the issue is how to make politicians and large corporations responsible for their actions, less greedy for profit, more transparent, without the ability to dominate and conquer other sovereign states and nations.