What does the future of AI hold for us? There are really an endless number of possibilities that come with the ability to create intelligent beings. Incredible advancements have already been made in the field of natural language processing and progress on vision is increasing as well. Most experts in the field, like Stuart Russell (Professor of Computer Science at UC Berkeley and co-author of Artificial Intelligence: A Modern Approach) and James Hendler (Professor of Computer Science at University of Maryland), predict a fairly common household presence of humanoid robots capable of a variety of tasks. Some however, have a more dismal look towards the future, like Professor Hugo de Garis of Utah State University who is doing research on and attempting to create artificial brains. Despite his desire to succeed in this field, in his on-line book The Artilect War, de Garis expresses his fear of what could come of his work.
de Garis feels that at the rate progress is being made in engineering fields, “artilects” (a term he uses to represent “godlike massively intelligent machine”) will take over our lives and potentially wipe humans out. He uses a few promising fields as evidence for this prediction, the first being nanotechnology. In reference to its ability to explore and analyze things on the tiniest of levels, de Garis feels that this technology will lead to extremely detailed information on the workings of the mind and brain, therefore allowing for flawless intelligence for artilects. Another concept de Garis mentions is Moore’s Law, with states that capacity of circuits(which consists of speed and density of transistors and such) doubles every 18 months. So far this theory has yet to fail and de Garis believes this trend will continue, creating extremely powerful artilects. Among other additional technologies he feels will lead to these unimaginably great machines, de Garis brings up artificial embryology. He believes that these techniques, when studied using nanotechnology, will allow for DNA equivalents in these artilects of which we will be able to manipulate, or more importantly, program the artilects to manipulate themselves.
Most people leave the scenarios de Garis describes to Hollywood, however, his predictions are quite unnerving. Where artificial intelligence will go from here no one can know for sure, but one can be sure that its historical impact is just beginning.