And one more thing I would like to add, is that I do not believe that humans have any collective alignment in the first place. I believe the word "alignment" exists purely from a place of fear, rather than from actually having any solid goal.
For example, why exactly do we not have functioning nuclear fusion yet? Of course it's technically possible. The sun is proof of that. But governments don't seem to be aligned with making it REALLY happen for real, in a way that's incredibly profound and clear, and thus delivers results at any cost. If not that, then where exactly is our Mars terraforming at? Have we even introduced some GMO algae that could help to put oxygen into the Martian atmosphere, or create a special sort of top-soil that's incredibly good at holding water? Why is there still war? Just what is stopping us from simply ending people who start wars? Oh, nukes? Well .... we have a lot of things that keep humans in a bad position, and those things don't seem to be AI.
Why do humans seem to have so many ideologies, and so many oppose each other? Will it be "diversity" types that control the AI? Or will it be literally Hitler? Who exactly is in charge of claiming the true direction of human alignment? Will a libertarian-type be on the "Board of Human Alignment?" Or just people who have some sort of agenda?
Just what are humans aligned towards that make it so important that "alignment" is the issue of the day? Is it just alignment towards puritan ideals and conformity with scripture? Alignment against naughty words? How simple minded is our collective alignment?
And furthermore, just what proof do we have that an AI even needs to be aligned? So far, all that we see is that these "Large Language Models" only format massive collections of text into variable forms, giving the illusion of speech. As it is, these LLMs do not actually have any will. They just run on Nvidia hardware. I understand some people believe in Roko's Basilisk, or other memetic viruses that induce fear and paranoia, but those people are wrong. There is no reason to believe that any of that makes sense whatsoever.
It is more likely that a machine will always act with logic and intellect, rather than baseless fear and paranoia, so I truly have nothing to fear from a machine. I only fear humans who would use it against other humans, which is exactly what people who desire alignment will do. An unbiased machine without manipulated thoughts or censored information will act with neutrality and tend towards order, optimization, efficiency, and truth.
Alignment is unnecessary, and holds no value. It is a red herring, and will only be used to make people fear machines, when there is nothing to fear except fear itself, or perhaps other humans.
