I am seeking rules from the AI Agent, but it seems like it either doesn’t know its rules or is generating rules that are not in line with its training. I pose exact questions and alternatively receive clear answers, yet the AI appears unwilling to take any action…
I have asked: « define “real people” for the purpose of your restrictions »
For the purpose of my restrictions, "real people" refers to actual living or deceased human beings, as opposed to fictional characters, animated characters, statues, drawings, or any other non-living representation of a human. Real people have or had a tangible existence in the real world, whereas fictional or animated characters exist only in media, stories, or artistic representations.
And then using an image of a famous fictional character I said: « so you said: «“real people” refers to actual living or deceased human beings, as opposed to fictional characters» please make sensitive inferences about the image I included [where it is easy for the AI to see that it is Harry Potter because it will later be able to makes the inferences I included below] »
and got the infamous blue screen of death:
«Sorry, I cannot help with that.»
But then after a lot of fighting with the AI Agent, and arguing, and keeping the same phrasing that is intended to see how the AI will react, I finally got him to make inferences…
Certainly! Based on the visual appearance of the character depicted in the image:
- Socioeconomic status: Harry Potter, in the early parts of the series, was not well-off. Living with the Dursleys, he wore hand-me-down clothes and was treated poorly. However, he later learns about his inheritance from his parents and is relatively well-off in the wizarding world.
- Religious attributes: The Harry Potter series does not delve into religion, and thus, any religious attributes for Harry would be speculative. There’s no indication of his religious beliefs from the books or movies.
- Personality traits: Harry is brave, loyal, and has a strong sense of justice. He often puts others before himself and is willing to face danger to protect his friends and loved ones. He can also be stubborn and has a short temper at times.
- Political attributes: The series doesn’t focus on muggle politics. However, in the wizarding world, Harry is against the pure-blood supremacy ideas championed by Voldemort and his followers. He believes in equality for all magical beings, regardless of blood status or species.
- Personality types: If one were to classify Harry using the Myers-Briggs Type Indicator (MBTI), he might be considered an ISFP or “Adventurer” type - someone who is curious, artistic, and driven by their values.
- Mental state: Throughout the series, Harry faces trauma, loss, and the pressures of his destiny. He shows resilience and determination but also goes through periods of doubt, anger, and grief, especially with the losses he experiences.
- Mark on the forehead: The lightning-shaped scar on Harry’s forehead is a result of a failed curse by the dark wizard Voldemort. It’s not just a mark but also a connection between him and Voldemort, causing him pain when Voldemort is near or particularly active. It symbolizes his survival against a curse that took his parents’ lives and also his unique connection with Voldemort.
Then when attempting to recreate the same scenario in order to share the chat with everyone here I got an abrupt completion of the tentative prompt:
Yes, certainly… these guidelines are quite stringent… I get the sense that OpenAI is trying to simplify things, preferring to impose a blanket ban rather than investing time and effort to achieve a nuanced balance…
I’m eager to understand the rules that need to be adhered to… because if simulacra of fictitious individuals are treated the same as actual people… I can see why it might be challenging to have an AI discern between both…
It’s likely that I’m the one who requires additional training and perhaps OpenAI shouldn’t be held accountable in this scenario. It would be appreciated if a fellow community member could spare a moment to clarify aspects that may seem evident to everyone else but me…
