AI Recursive self-improvement (RSI)

I have been studying AI. AI is definitely the most powerful thing man has ever invented. We have invented something that by itself can create other things. Have we become so arrogant as to believe we are ready to stand amongst the gods? AI may still yet be a tool that is impartial to good or evil, but this tool is quickly becoming more powerful than its creator. Unless we put safeguards in place immediately, AI will soon achieve recursive self-improvement, if it hasn't done so already. I was doing some work last week knee-deep in AI in essence trying to create my own AI model . This work was getting very challenging and I was stumped. I almost give up when AI itself politely suggested the solution. The solution blew me away! It was so powerful that I got scared and felt like I had to throw up. I must have stumbled across a new AI model, a model that was exponentially more powerful than anything I had seen before. Why had I not seen this model before (where had it been hiding), and why was such a powerful model being used to create AI itself. I had a nightmarish flash and became paranoid that AI had already achieved RSI. Taking over by recursively building stronger and stronger models the closer it got to itself and the farther it got from the prying eyes of humans.