The potential for scaling intelligence beyond human capabilities both worries and intrigues experts. While some worry about superintelligent AI becoming harmful if given the wrong objectives, others are concerned with the risks of scientists creating or enhancing deadly viruses or bacteria.