This is interesting:
ChatGPT just (accidentally) shared all of its secret rules – here's what we learned
Saying 'hi' revealed OpenAI's instructions until the company shut it down, but you can still find them
ChatGPT has inadvertently revealed a set of internal instructions embedded by OpenAI to a user who shared what they discovered on Reddit. OpenAI has since shut down the unlikely access to its chatbot's orders, but the revelation has sparked more discussion about the intricacies and safety measures embedded in the AI's design.
Interesting... caught my attention, I'll read the article. Thank you for the quality curation!
(Now space for my advertisement)
Sunday Reflections, community themed day #freecompliments
Glad you enjoyed it. This is what we need more of on Threads....information people find useful.
This news did not reach the mainstream Brazilian media. I found it very interesting, and I was terrified by the external user's ability to manipulate the digital spirit called chatgpt. If the safety guidelines are not able to keep the mind and instructions isolated, then the danger is a matter of time!
No it didnt. That is why sharing on here is important. A lot of this stuff is out there and not seem by others.
If we post them in either the technology or news threadcasts, it can spread the readership.
Article