Feel free to leave a comment.
The whole premise, while an interesting hook, is kind of pointless when looked at factually - the reason why people speak to LLMs "like a human" isn't because of a complex relationship on how we treat others, but because the LLM was trained on human writing, and performs best with appropriately human-like inputs.
I wanted to see more insight into the tech itself, but a large part of the article just talks about what bad things "Silicon Valley" (a single, monolithic entity) has done in the past. It ignores both that generative AI is a new frontier that has completely different challenges (including companies being unable to "steer" outputs as much as a recommendation algorithm), as well as the sheer existence of any free and open AI software.
I understand your point about the reason we tell people to speak to LLMs "like a human" has to do with functionality, but I don't think you can discount the push towards viewing these systems as human-like and more relatable, especially in the marketing and product applications (AI companions, customer service help et. al) both from large and small scale tech platforms. And regardless of whether the functionality requires you to interface with it like a human, doing so still has an non-negligible effect on you.
Also, when I refer to Silicon Valley, I am more so talking about the pervasive thinking and practices that are common and codified in the industry. The thing about the "bad things" you mention is not just that they were done in the past(they still happen), but that the underlying motivations that led to them in the first place still exist. Yes, Generative AI is a new frontier, but it is not immune from the motivations that have been present in the industry for decades.
I agree that I could have done better with regard to offering more insight about the tech itself, but I didn't want to write a technical article, I wanted something more personal.
Anyway, thanks for the read, and I appreciate the feedback.
