The X @Safety team should launch a new initiative aimed at identifying unwell people who argue with Grok as if it's a human or portray other traits of AI psychosis, so that they can direct those folks to seek the help of a mental health professional.(Jameson Lopp)