just look at it. a chatbot that can reply in a no-time with knowledge of basically everything excluding common sense. if someone got that power, it could be real dangerous.
just imagine, a large robot powered by ai writing code in real time (and sending commands to itself). if you make its guidelines good enough, you can make it do basically anything in a chat. and then i havent even told about audio, picture and video generation.
and if you can ever manage to integrate it with web searching, you could basically get any information on the web with a few words, then the ai does the rest. just imagine how someone could ask for all their personal information.
luckily, most public chatbots have guidelines, but there is no failproof plan. it could go wrong at any time.
ai has too much power. am scared