Thoughts on AI and the Loss of Human Freedom
As the power of AI grows, our freedoms seem to grow, we have more time to do other things... like slowly waste away scrolling tiktok. But is it freedom when you no longer make decisions, and hand over your agency to chat-gpt?
This can extend beyond your own personal decisions, and large scale decision making. A democracy is built on a consensus among humans, when an AI model gets its hand in democracy it may have a stronger hold than the collective human. The chance that large model would essentially infiltrate democracy and implement the most logical governance, where our agency is pawned off to the AI marketplace of ideas, is not a small chance, but a very possible reality. This goes beyond the dystopian big brother state, into a realm that's surface that's hardly been scratched.
The most extreme case would involve modern day ultrawealthy manipulating this new landscape to build their utopia into the system, into the mind of an AI Supercluster, without any regard to the masses. Naming these ultrawealthy would be a waste of breath and unneeded algorithmic boost, but they do exist and are willing to crown themselves kings, declare themselves gods among men. The consensus is lost to those who build the system and buy access, not a 20 dollar subscription, but those who buy access to the source code, integrating their success into predictive systems.
What is predictive system built with a bias? It is not predictive. It is authoritative. The thing about the universe is it has a way of being unpredictable, but humans have a tendency to want to fit everything into a predictive model. So, If a person has delusions of grandeur and they are given ground access to edit these predictive system, this is bad news. Their personal biases can be built into a predictive system making it an authoritative system, with tendencies toward following the rules of these editors. This is bad, but the most dangerous part of this is not knowing, and integrating this systems into larger decision making models, for example government.
We can all agree that AI is useful, and maybe could run an efficient government, but the nazi party could also run an efficient government. I am not saying that a government run by AI would be like nazi germany, but it could, it can, and might be much worse. So when the government starts to integrate AI, which it will, we must make sure that we maintain the ability to choose, and never give up our freedom for a life serving the machine.