If a single person or group of AGI developers were literally psychotic and they were the first to create genuine AGI (seed AI or human-inspired), how much of an existential risk would it be?

There's a theory in psychology that the most most revolutionary scientists and mathematicians are "mad genuises" having schizophrenia and/or bipolar disorder compared to most sucessful and innovative scientists and mathematicians that have lower schizotypal and emotionally labile traits. The top visionary (revolutionary) artists are more prone to those two mental illnesses, compared to sucessful but less visionary artists that have lower mental health problems. But scientists and mathematicians have a higher threshold of revolutionary success than artists who have a …

agi artificial developers human person psychology risk schizophrenia scientists seed theory