Advertisement Article continues below this ad While there has been substantial research on methods for keeping AI from causing harm by avoiding such damaging statements – called AI alignment – this ...