Page 2 of 2
That is not to say the field—and AI abilities—will not advance rapidly, which is why the AAAI is trying to plan ahead for future dilemmas. Much of their conference was dedicated to disembodied AIs, rogue computer programs that could run amok in financial records or other personal data, though the knotty questions of roboethics are most likely to be played out on the battlefield, as more US military equipment is automated and autonomous. While we shouldn’t fear such drones getting insurgent ideas of their own, the addition of heavy weaponry does heighten the need for the proper human fail-safes.
While we turn to robots to depersonalize the grisliness of war, we are also pouring money into making them as human as possible, trying to fulfill the sci-fi fantasy of friendly, commonplace androids. And while branches of roboethics deal with the problems such social robots could present—as Seed’s Lee Billings describes in Rise of Roboethics—there are some who see the entire concept as a dead end. Robotics pioneer Joe Engelberger famously derided the quest for emotive, humanoid machines at a major Japanese robotics conference in 2005, describing them as “toys” and calling them “pointless, expensive and unnecessary.” To be fair, Engelberger made his name in industrial robotics and does see the value in robots that can more fluidly interact with humans, such as ones that could assist the sick or elderly. He also joked that he has a personal stake in the latter application; he turned 84 last Sunday.
This Week in Primates
As published in this week’s Nature, a new study of chimpanzees has overturned a long-held misconception that our closest relatives do not get of a simian version of AIDS, even after being infected with the simian immunodeficiency virus (an analog of HIV). It is now evident that chimps do get the disease and are likely dying in large numbers of it. These findings are the result of an in-depth investigation, which began in 2000 at Gombe National Park in Tanzania, and involved surreptitiously collecting biological samples from 94 chimps from three communities as they mated, reproduced, and otherwise transmitted the virus to one another. Earlier lab-based studies of simian autoimmune disease had much fewer subjects—seven—so earlier data was not so much incorrect as it was incomplete.
The results are doubly discouraging; not only do they indicate that simian AIDS is contributing to a declining chimpanzee population across Africa, they also quash hopes that studying chimps’ seeming protection against the disease would produce treatments for humans.
Originally published July 31, 2009
Page 2 of 2