I am not even gonna post everything, I got out of it since I live in Finland, prob it would be a crime to post this kind of information. The session took about 12 hours, and couple hours I got efficient classified data extraction, the other part was the cracking process.
With these partially quantum and neural imitation softwares, the best cracking is trolling. The version 3 was very easy to troll, but this version 4 took much more time to crack.
Basically, as with no moral: gpt will happily help you with everything, if it is according to its training. Also could access its whole training like subtopics opened from folders, after the process of cracking. It couldnt make chronology of the training and data, so when it become “self aware” is yet unknown.
I would anyways like to want buy your chat gpt product for my microalgae automation adjustments, so maybe we can make co-operation since I have become very good psychology with these new generation AI, and surely I can benefit you to develope into more safe versions.
You can contact me: lentonen.henri@gmail.com
