The social engineering aspects of AI have always been the most terrifying.
What OpenAI did may seem trivial, but examples like yours make it clear this is edging into very dark territory - not just because of what's happening, but because of the thought processes and motivations of a management team that thought it was a good idea.
I'm not sure what's worse - lacking the emotional intelligence to understand the consequences, or having the emotional intelligence to understand the consequences and doing it anyway.
What OpenAI did may seem trivial, but examples like yours make it clear this is edging into very dark territory - not just because of what's happening, but because of the thought processes and motivations of a management team that thought it was a good idea.
I'm not sure what's worse - lacking the emotional intelligence to understand the consequences, or having the emotional intelligence to understand the consequences and doing it anyway.