Futurography readers share their opinions about killer artificial intelligence.

Some readers were skeptical of the idea that A.I.'s interests could ever correspond with our own, despite the work of researchers such as Stuart Russell who are trying to ensure that computers can learn to recognize what's most important to humans. As one who took such a position put it, "A.I. will likely just be a very, very smart machine and the concept of'interests' might not bear." Instead, "We should worry more about whether the interests of private and military A.I. R&D teams align with our public interests," a reader suggested. Another wrote that "terrorist groups and … enemy nation states" presented the greater risk. And others continued to hold that the real trouble is that there's no such thing as human interest per se, as did one who wrote, "We humans can't even agree on what constitutes'good' and what constitutes'evil.' "