Marvin, Douglas Adam’s deeply depressed android creation was given what some might consider to be a rather cruel gift… not only was he bestowed with unfathomable intelligence, he was also the recipient of a gamut of human emotions. This combination of smarts and feelings, mixed with the menial jobs often tasked to him by the rouge president of the universe Zaphod Beeblebrox led Marvin into a permanent and unrelenting state of despair. Now you might think, surely a bored, depressed robot is just the stuff of stories. But, it might just be time to think again, autonomous robots and Artificial Intelligence is no longer the stuff of the far future or simply flashy action film fodder, no, this stuff is starting to get real.
The UN recently staged a conference focused on discussing the use of autonomous killer robots and determining whether the use of such technology could potentially violate human rights and the laws of war. Director of the Ethics and Emerging sciences group at California Polytechnic State University, Patrick Lin, has weighed in on these deliberations, pointing out, in this article, an inherent need to ‘personify’ robotics, questioning whether human standards should really be applied to objects that don’t have any capacity for moral concern. We might dismiss this as a problem for future humans to deal with, but the reality of AI is fast approaching, and it has me wondering, where do these ‘human’ expectations for robots and AI come from?
Where else to look but that enduring fictional genre… science fiction? Science fiction (aka Scifi) has long used the cutting edge of science and engineering as fodder for the imagined futures of humanity. Within this realm of imagined futures, the development and evolution of Artificial Intelligence (AI) stands out as a Scifi favourite. From Isaac Asimov to Spike Jonze and Douglas Adams in between, it’s hard to find an example of a fictional future that doesn’t contain human-like robotics and AI. It seems to me that there is a very strong link between these historical, fictional visions for Humanity’s future alongside autonomous robots, and the contemporary, cutting edge of robotics.
Across the world there are a number of organisations working to develop artificially intelligent, human-like robots. Hanson Robotics in the USA have built a series of robots that have been designed to appear as human as possible, from their ‘frubber’ skin and expressive faces right through to their ‘personalities’ the resemblance is a little eerie at times. One of Hanson Robotic’s most prominent creations is BINA48, a robotic bust that has been bestowed with the looks, memories, ideas, feelings and beliefs of a real individual, a woman named Bina Apsen. By giving BINA48 all of these ‘human’ characteristics and information, the hope is that ‘she’ might be able to evolve her own intelligent consciousness. Another company, Hiroshi Ishiguro Laboratories have a similar approach to the development of human-like AI. Their robots are designed to appear like real humans, they also have two models, the Geminoid HI-4, which has been developed to have a larger degree of autonomous ‘behaviour’ and the Geminoid F (a robot I had the pleasure of ‘meeting’ back in 2013), a ‘lite’ version of the Geminoid HI-4 that has fewer autonomous capabilities. A third example of an attempt to create a hyper-real, human-esque robot is apparent in Akio. Designed by robotics enthusiast Dr Le Trung, Akio is one man’s attempt to create a human-like robotic companion to assist individuals with domestic tasks. Akio’s has been given the appearance of a young woman, however, has not be explicitly based on a particular human individual.
All of these robots, developed in a bid to create an autonomous intelligence, have been given human attributes, both in their form and with the information that is intended to define their ‘personalities’. I can’t help but think that these attempts to recreate human intelligence stem not only from our inherent narcissism as humans, but also from what we consistently see in science fiction. In every story that I can think of artificially intelligent robotic characters almost always take on human characteristics (no matter how flawed they might be). Consider Ultron, the main antagonist of Marvel’s latest blockbuster generates a machine-made body that looks distinctly human, or the aforementioned, perennially depressed android Marvin. Similarly, Samantha, the nearly omnipresent operating system in ‘Her’, does something extremely human when she falls in ‘love’ (even if it was with hundreds of different people all at once).
Speaking of ‘Her’, it seems that a ‘Her’-like world is set to occur in a not too distant future. Google is in the process of developing new technology that could enable computers to develop ‘common sense’ within a period as short as a decade. That’s right, soon we might be talking to our computers as though they are our friends… I wonder how that will change the current social isolation that seems to be developing out of our Internet oriented societies? Rather than being a device that allows you to remotely connect with your friends, your computers and phones can become your friends instead…
It seems that there are clear connections between what has been presented in science fiction and what we are now seeing at the cutting edge of technological development, but there’s also a chance that this obsession with artificially emulating human intelligence doesn’t derive from scifi at all and maybe it’s all in the spirit of competition, with these developments entirely focused on developing AI that is capable of passing the Turing test. This test, founded on the work of mathematician Alan Turing, has been developed to determine whether a machine is ‘intelligent’, based on its ability to ‘trick’ a human ‘judge’ into thinking a machine is also human. Add some money into the pot (there are a number of prizes on offer to anyone who can create such intelligent programming) and you’ve got even more motivation to focus on the creation of human-like artificial intelligence.
It seems that we are incapable of separating intelligence from humanity. We cannot envision that any other intelligent, autonomous ‘being’ could have intelligence or abilities that are not rooted in humanity. And, as we see in the aforementioned discussions on the use of autonomous robots in war, that is, we are starting to apply human standards, morals and ethical expectations to machines. Perhaps we need to look elsewhere for inspiration when it comes to the development of AI, to create an intelligence that is something other than Humans 2.0…. what do you think?
Image: Hannah Greethead