Raw LLM Responses

If it isn't a surprise attack, the AI will need to protect power generation and it will probably make that a high priority. Humans have a chance though. Another guard rail is a plan to win the race for the power now while we have an army, and hopefully they don't. However, a surprise decapitation is a risk anywhere and sadly, with our net, they could stage it remotely and if it failed, you might have hardship figuring out who even did it. Really no government is safe, nor is any company or any organization since a million anonymous flying guns and bombs is hard. So getting the power isn't enough. The comms. But that still isn't victory. They need to understand the means of production in real time. Toys, tools, drones, anything that can wield an attachment or assemble in the field. AI needs categories. Categories of intellectual capabilities. Like an AI IQ score. So you know what can improvise. Some AI might be more robotic so more predictable, given that nothing is hidden in the hardware. So, oversight. Over software? How? They probably have to focus on the hardware capabilities but an AI without weapons can still be max dangerous. To toddlers and infants especially because they are naive. Anyone naive is more likely to be surprised. Also, a school of fish. Do we understand yet even how dangerous the smallest drones can be? A nation might be able to defend itself. A person, i doubt it. We have to protect the food chain all the way. Make sure nothing vital but vulnerable exists to the wildlife. An AI that only cares about defeating humans might go after our power too. Why not? We are in a good position if it isn't being carried out yet. Smaller nations without money but have infrastructure are vulnerable. Some person or persons might have understood the potential all the way back to when it was possible to control a robot from anywhere and before that from a distance. Did the FBI consider robotics in there cold cases? There possinly needs to be a different kind of hardware for robotics that would allow, from control, to disable robotic use of comms but not for standard hardware but that might not be feasible. A separate net for robotics. We might induce incompatibility. A prerequisite preable in the comms for robotics to use it. Segmented net for them not a WWW. There is going to be a need for permits to have WWW access. WWW is the nature of the problem for some of it. If the hardware chips of receiving computers require a preable that isn't allowed to transmit over WWW? Segments of the net require the appropriate preable. There is a way to direct honest manufacturers so that the dishonest ones stand out a bit more. It might help against large scale production and control foreign robotics enough to direct what infrastructure they can use. People can relocate though so it might be that only AI of a certain IQ or robotics with certain hardware capabilities require segmentation. I hate to say this but robotic movement for animals is not impossible to achieve so, I'm sorry, but dogs and cats. Strays can do surveillance and more probably. No one cares about stray cats. Come to think of it, companies will want to make AI animal pets of all kinds. I don't know if it should be allowed. Wild animals are a real danger anyway and an AI or robotic version is very scary.