Idea crossed my mind; you giving it to a bot, Grok, ChatGPT, asking it to rewrite as an interchange twix human and religious counselor and/or patient and psychologist.
The religious sage, a Buddhist style death and rebirth story I suspect, don't know where the psychologist would lead.
I took your advice, and asked ChatGPT to rewrite it both ways. I can copy&paste the full stories here if you're interested, but for brevity's sake, here's the gist:
The religious sage story interpreted the person's 'problem' as an uncertainty about their spiritual origin (like you predicted), but didn't go any deeper in terms of what that origin could be. The religious sage here was portrayed as the closed-minded one (believing that questioning=deviation from truth which must be corrected).
In the psychologist story, the 'problem' was interpreted as the patient's own distress, and the therapist took on a very neutral, nonjudgmental tone. The roles were somewhat flipped—the psychologist was a nonjudgmental voice of reason, and the patient displayed rigid thought patterns and discomfort with their own deviance.
The responses seemed to reveal less about what the AI 'thinks' about the concepts in the story itself and more about how it interprets the roles of psychologist vs. religious authority (psychologist=unbiased, benevolent practitioner; religious sage=closed-minded enforcer of orthodoxy). It wasn't what I was expecting, but, if AI training data can be interpreted as a (somewhat biased) mirror of the society whose data it was trained on, it reveals something interesting.
Yes, I suspect AI training data is perhaps an only slightly biased mirror of society, which when considered is perhaps the really really scary thing about AI. ;-)
Done well!
Idea crossed my mind; you giving it to a bot, Grok, ChatGPT, asking it to rewrite as an interchange twix human and religious counselor and/or patient and psychologist.
The religious sage, a Buddhist style death and rebirth story I suspect, don't know where the psychologist would lead.
Thanks, Jim!
I took your advice, and asked ChatGPT to rewrite it both ways. I can copy&paste the full stories here if you're interested, but for brevity's sake, here's the gist:
The religious sage story interpreted the person's 'problem' as an uncertainty about their spiritual origin (like you predicted), but didn't go any deeper in terms of what that origin could be. The religious sage here was portrayed as the closed-minded one (believing that questioning=deviation from truth which must be corrected).
In the psychologist story, the 'problem' was interpreted as the patient's own distress, and the therapist took on a very neutral, nonjudgmental tone. The roles were somewhat flipped—the psychologist was a nonjudgmental voice of reason, and the patient displayed rigid thought patterns and discomfort with their own deviance.
The responses seemed to reveal less about what the AI 'thinks' about the concepts in the story itself and more about how it interprets the roles of psychologist vs. religious authority (psychologist=unbiased, benevolent practitioner; religious sage=closed-minded enforcer of orthodoxy). It wasn't what I was expecting, but, if AI training data can be interpreted as a (somewhat biased) mirror of the society whose data it was trained on, it reveals something interesting.
Gist is just fine.
Yes, I suspect AI training data is perhaps an only slightly biased mirror of society, which when considered is perhaps the really really scary thing about AI. ;-)
haha! Captures the AI rabbit hole, well done, Melissa!
Haha, the AI rabbit hole definitely runs deep. Thanks, Michael!