There has been a lot of talk about AI escaping human control. I have not heard any discussion of who should control AI. AI advocates are trying to push the idea that AI will be developed for the betterment of all of humanity. The problem with this idea is that the billionaires who are behind the development of the most powerful AI models are not proponents of altruism, generosity or equality.
The level of access to AI abilities for the owners of these companies will not be the same as the access that is granted to the general public. When talking about how AI will kill us all within ten years, they talk about AI doing things like hacking into nuclear weapons controls. There has been no discussion of what would happen if the owners of the AI models manage to stay in control and deliberately use them to gain control of nuclear weapons, taking them out of the control of elected governments.
Another concern is the use of AI for biological weapons. What would happen if the far-right agenda of white supremacy is acted upon by the owners of AI to develop a biological weapon that targets specific ethnic groups by using biological markers inherited from specific geographic areas? The elimination of funding through USAID, resulting in the deaths of large numbers of people, could potentially be only the first step.
I am using this as an example, not because I believe this is what will happen, but because we need to consider how this technology could be misused, not simply by individual users, but by the owners. Misuse of AI technology would not be limited by traditional borders, but by the domain of the ever-expanding internet, which can now be accessed in some of the most remote places on the planet.
Where AI legislation is passed, how can it be given teeth that are of consequence to billionaires? In a world where rich people are rarely held accountable, even to a fraction of the extent that they should be, which is in proportion to the degree of harm and the number of people harmed, how can we have faith that attempts to legislate the creation and use of AI would ever be enforced? Whether AI is under human control, or its own, I doubt any of us are prepared for the consequences of unregulated AI development.