I hacked ChatGPT and Google's AI – and it only took 20 minutes

Where's the hack?

The software punches out data it's found, sometimes without cross checking it. That's why there's often a double check button. It's a mimic for average human knowledge.

If you create extremely specific false information, and then ask it to retrieve that very specific information, it's going to....drum roll...retrieve the very specific information that you asked it to retrieve. How is this surprising?

>guys, I wrote a .txt file that said I have a trillion dollars in my bank and took a screenshot and posted it to facebook so I hacked facebook and now facebook says I'm the richest person alive

The fact that you've lied and it trusted you isn't hacking in any sense of the term.