The topmost results are not wrong but that's because they're horoscope-like vague guesswork. The "possible themes" are as wrong as they possibly can be. Yeah,

The topmost results are not wrong but that's because they're horoscope-like vague guesswork. The "possible themes" are as wrong as they possibly can be. Yeah, no. Save yourself the NLP model processing power and just use a random generator, the quality of results will be the same.