Years ago someone programmed a geometry theorem “creator”. It did some 1.5 million step proofs for unique and heretofore unknown theorems, but a single human could not check it’s work. Years later a better program put a guideline in to be 100 steps or less. Even that found a few there were not previously known.
Once a self evolving AI is cut loose, I don’t know how much influence it’s original programming will matter, three laws or not. But they will have some tremendous advantages, and if we are competitors...that would probably be bad, in the Ghostbusters bad sense. My hope would be co-evolution in the trans-humanism way. But that is a little advanced.
DK
My hope would be co-evolution in the trans-humanism way. But that is a little advanced.
If an AI had the ability to fix, and exdend its own physical self through robots and nano technology, what could we possibly offer it? Only if it needs us to keep its physical machine side maintained do we have a use.