Pinocchio Syndrome: AI, Robots, and Fear

Created by Yul Jorgensen aka deviantart artist FATBM

By Lon Maxwell, Reference Department

I and most people in their thirties and forties grew up in an interesting time when it comes to robots and Machine intelligence. Robots are the automated devices that build our manufactured goods, the programmed tools that weld, test, and gauge everything we use. They may even for a few lucky ones have been something we got to play with as kids in school robotics labs, programming multi jointed limbs and pincer grips to move small objects from one side of the table to the other. For me, they were the first peek into Machine language. This was not, however the only way I saw robots. They were also the shining metal sidekicks, the implacable companions, and often the comic relief of some of my favorite books, movies and television shows. I could go from watching a sci-fi film on Friday to playing with an old Unimate PUMA at the local community college’s Saturday kids program the next morning. I almost didn’t see them as even similar versions of the same thing. Now that I’m an adult however I can look back and see my old Saturday morning friend was the Homo Habilis to today’s Homo Erectus. We are still in the early days of robotic evolution, but the pace is quickening and the robot who wants to be a “real boy” could be just around the corner.

IBM 7090 computer and personnel from 1961

You can’t take a look forward and appreciate the amount of journey ahead without knowing how far you’ve come. True electronic computers go back to the old vacuum tube run models that took up the same space as an entire gymnasium and were programmed with punch cards. Then came central mainframes with dumb terminals, mini systems and finally the personal computer we know and love still. In that time we migrated through magnetic tape and cartridges, through the innumerable floppy discs and hard discs to the thumb and hard drives of today. The robots have changed a bit as well. Aside from the science fiction progression we’ve seen from the Lost in Space robot to…well…um… the Lost in Space robot (1965 and 2018 respectively) there has been a huge swing from the early mechanical arms to the more modern, and yet still 18-year-old, Asimo.

Pepper is a semi-humanoid robot, manufactured by SoftBank Robotics, designed to read emotions

The future of robots is heading down two main tracks, Humaniform[1] and non-traditional robots. The Humaniform robots are learning (more on that in bit) to respond to human emotion. In some cases they are working on mimicking facial expressions with plastic and servos. It is among this variety of robot that scientists are working on human like movement. The non-traditional types are just as surprising. Military vehicle=s and drones are where the largest strides are coming from. Self-guided autonomous devices are delivering supplies and personnel. Dangerous missions are being performed with pinpoint accuracy by computers that are learning to modify their tactics. They might not look like C3P0 but the military machines are where we see science fiction meeting reality. Let’s just hope it’s more Bicentennial Man and less Skynet.

Artificial, or machine, intelligence is a different story. Until today the general maxim for working with computers is that they do not make mistakes. They do only what we have programmed them to do. Any errors are really the fault of the programmer. Going into the future we may see computers that are capable of extrapolating their own options and acting outside of initial human programming. This artificial cognition has a lot of people excited, good excited and bad excited. Advances in computing ability are allowing for computer scientists to attack the problem of A.I. in many different ways. They are approaching the problem from the angles of symbolic learning and human brain simulation among several others.  They are also using the latest tools, networks to simulate neural pathways and statistical models to build decision making. One big question in artificial intelligence is the ability of man to create a moral structure for the A.I. brain to exist within. Many do not believe we can safely create a friendly A.I. with the level of knowledge we currently possess. That might explain why fiction has a lot more HAL 9000 clones than Commander Data.

Good vs Bad… who will win?

The real future that many of us hope for is one that brings these two things, artificial intelligence and robotics, together to make a robotic being that will help us forward. Like many of the advances we have seen our imaginations are directed by the stories we hear and see and read. Advances in both fields will lead to the point where a breakthrough occurs, and it will be sooner rather than later. The only questions left to ask are will it be a benevolent discovery and are we worried about if we can rather than focusing on if we should.

[1] The terms humaniform and non-traditional are ones that I have chosen. Humaniform is a term I have lifted directly from the work of Isaac Asimov.

---

Further Reading (and since it’s Friday the 13th, have 13 books):

• I, Robot by Isaac Asimov (F ASIMOV)
• Do Androids Dream of Electric Sheep by Philip K. Dick (LP F DIC)
• Saturn’s Children by Charles Stross (eBook)
• The Alchemy of Stone by Ekaterina Sedia (eBook)
• Neuromancer by William Gibson (F GIB)
• Robopocalypse by Daniel H. Wilson (F WIL)
• “Silently and Very Fast” by Catherynne M. Valente (808.838762 MOR)
• The Automatic Detective by A. Lee Martinez (F MAR)
• The Windup Girl by Paolo Bacigalupi (F BAC)
• Hyperion by Dan Simmons (F SIM)
• The Moon is a Harsh Mistress by Robert A. Heinlein (F HEI)
• Otherland by Tad Williams (F WIL)
• Ancillary Justice by Ann Leckie (F LECKIE)