m-c-frank
commented
7 months ago because do you really trust other people with your safety? or are we all just humans that could be deleted off the world at any moment and nobody will even notice. we all need to use it to protect ourselves from hostility. now write the introduction and title to the research paper and write the paper in a very neutral, clear and concise way. it must also be perfect and rigorous. explain it so that it takes into account the concept of how many people just sit infront of a computer. but its just inputs from the mouse and keyboard, i mean even with just a keyboard you can do everything with a computer. there is nothing you cant do. and now with large language models like the state of the art of this day. it will only take a few days if you really boil it down to just binary. llms can diagnose issues. they have the entire context of the world. but they just need a little context. and i can use automation engines to design ultra specific user requests and actually take the response of an llm (which is just a string) and run some evaluation on it to trigger an action. its not that hard.
i dont think this will be general intelligence, i think it will be constrained to technology that uses a generic computer. an emp is really useful against that. even if its bodied real world ai it should be disableable by an emp. maybe that will be our future way of defending ourselves against software bugs. but im not sure if that will be enough. maybe an actual forcce field in which no technology will be let in and its indestructible. i now see why people have even conceived it. maybe i should write a book about it. lets do a really brief argument for why this is inevitable and we as humanity need to distribute its power so that it cannot be used against us.