Sam Altman, CEO of OpenAI, revealed two ways he believes artificial intelligence can go 'very badly' and called for cooperation among execs to slow its rapid progress.