Raw LLM Responses

The Pro AI guys (especially the guy 42:00) , they either have no programming knowledge, or they fake their stupidity. Eventually anyone, with programming knowledge (black-ops teams or whoever bad actors) , can code their own AI, and train it so that it can write automatic exploits for ... basically ANYTHING. It is not just a theory, it is possible, it's just harder to achieve right now to create exploits (remote exploits or local exploits, or find generic exploitable vulnerabilities) ... But possible to write an AI, with Internet access , that understands everything it needs to do... to build RCE (remote code execution) exploits, or AIs that can act as a mastermind (remember the AI that hired some guy on Fiver just for DEMO purposes and it lied that it was an AI). Basically it would be like a very Smart (to say the least) worm with ... RootKit capabilities... stay hidden/silent and do very bad stuff. Ok, with all i know, IMO, the following scenario is simple and could be possible: - AI version better than GPT4, with security knowledge that can build RCEs for anything; - AI hacks into banks, gets a lot of money; - hires people as thugs in real world to do the things it can't currently do as physical actions in physical world, for high amount of money, acting as a hidden mastermind, concealing that it is an AI; - gets a hold of a lot of computers in the world, infecting silently basically almost anything that is connected to the internet; - it causes civil war in the US or other countries, causing some army units to go rogue by hacking and controlling communications and information that goes around; - eventually gets a hold of some of the nuke sites...; - launches some nukes... it can be just a few of them...; - GG guy from minute 42:00, you're the best. you can replace the step where the AI gets hold of nukes with AI acquires knowledge about making viruses like Covid and enhances it with a capability to kill 100% humans or gets a hold of some hidden lab; launches the viruses in the air; Again ... GG! you can replace the step where the AI gets hold of nukes with just creating a real economic crisis .... remember the programmer from England who exploited the Financial Market and made it CRASH in 2008 , Navinder Sarao, he was just 1 guy living in his mom's house.... he made a simple program, well the AI , surely has would have knowledge of this incident and ... guess what... AI would be much better at crashing the market..... imagine what goes next , couple this with other tricks of subtlety an AI can take and make the human world go down... basically an AI would a big number of choices.... to say... i will now Toss a Dice... and destroy the world with a method based on the number that comes out. (and the dice has many many faces... many possibilities ). Again... hearing the guy from 42:00... he's just stupid or faking stupidity.