One thing that academic (not for profit) ai researchers understand is that computers cannot make management decisions, as a computer cannot take responsibility when things go wrong the way a person can. This is already an issue for traffic violations for “self driving” cars piloted by foreign labor overseas.