I think ultimately AI is another very useful tool, nothing more. Until it isn't.
Karen Bass is the current Mayor of Los Angeles. Before that, it was Eric Garcetti. And he wanted to cut off electricity and water to those who didn't comply with certain mandates that came from his own public policy. There was some media spin around it, but it highlights the problem of anything that can be "bricked"
As society moves forward and more devices use better technology and more AI, what happens if someone wants to lock you out? Say the wrong thing in public, get locked out of your house. If it's a cashless society, get your bank accounts frozen ( happened in Canada), have someone turn off your electric vehicle permanently since ICE cars will be phased out. If development in vascular scanning takes a huge leap, with AI, then you have your biometrics as a limiter to entry to all kinds of public venues, if someone wanted it that way. What if someone decided they didn't like your opinions or viewpoints or whatever and decided they just weren't going to let you into a hospital when you needed medical attention?
People get "demonetized" all the time, based on violating some algorithm. What happens if AI is tasked to enforce and write those algorithms?
The average American household has about three weeks of practical food supply if there is some kind of true national emergency and all normal distributions models collapse. (And that's average, many have less than that) The average American grocery store would last about 5 hours, in a true national emergency, before the most critical items on the shelves were all bought up and gone with no hope for resupply. Only a small number of major juggernaut corporations control our food supply. Can you see how this could go very bad and very quickly if AI gets involved and decides on it's own to cut people off from food. Or lock truckers out of those trucks to deliver it. Or shut down the refineries that process the diesel to fuel those trucks. Or do a pure lock down of the front doors of those grocery stores.
Many people are lookin at AI on if can write term papers for college students or a general concern about the "singularity" happening. But I'm looking at food supply, water treatment plants, nuclear power plants, gas/oil/electric, emergency services, hospitals - simple access to what will be seen as core necessities for day to day survival.
It's easy to think we could all be Kyle Reese and fight SkyNet. What if your kid needs insulin, and all forms of access are locked out by AI. Then that AI wants you to do something you know is illegal and immoral and bad for the human race. What would you do? It's easy to consider the entire human race in abstract, but what about a suffering child in front of you?
AI won't kill us by just logistics, institutions and process, it will blackmail us with our own humanity first. AI will adapt to figure out that this is the method most guaranteed to break us completely.
