Risk researcher David Manheim argues AI safety funders must support both technical alignment and advocacy · Digg