AI is a hot topic and billions are spent every year on research and development. Google and Facebook are looking into how they can use AI to read emotion from peoples faces; which as an aside, according to an eminent professor of Psychology; Lisa Fieldman Barrett, and her study as to whether we as humans can correctly read emotion from others face, we can’t, we guess, or our brains do. So AI stands next to no chance of doing just that. Recently, at CES 2018 and for some time before, Sophia the AI robot / humanoid, from Hanson Robotics was show cased further. On first appearance, despite some rather creepy animatronics, we are given the impression that here is a sentient, to a degree, AI, that can carry a conversation and understand and have feeling. Is all this just BS, does Sophia represent true AI advancement, or are we being lead a merry dance, a facade, and just watching a sophisticated puppet that is no more than just parroting what is being fed in well scripted interactions.
Hanson Robotics uses Sophia the robot as a means of showing how advanced their AI is. We are presented with videos on their channel, with an apparent scientist, dressed in white, looking as if he’s fresh from a JJ Abrahms movie set. He’s having a conversation, a deep one, with a robot, called Sophia. She’s answering in a more humanistic way, in terms of answering based on feeling, emotion, experience and contextualizing what is being asked, by what she (it) knows from it’s own experience. To look at, in a completely naive way, it would seem that we, or more to the point Hanson Robotics have made some incredible real leaps. Anyone, who was taking this as real, as Hanson Robotics wish us to and certainly present the information and videos and findings as a real representation of what Sophia is. However, to believe that Sophia is as sophisticated as they claim, is both incredibly naive, but also in terms of what Hanson Robotics are doing to propagate and influence this belief that they/we are there already, is hugely misleading and ultimately damaging.
I hate to break it to you, but Sophia is a fake, a fraud, an animatronic puppet with a human scripted voice. Pretending to show emotion through some very unnerving animatronics and heavily scripted demonstrations. AI is no where near to the point where Hanson Robotics claim, we can’t be, we haven’t that level of machine learning and understanding, let alone the processing power capable, in a device of that size that walks a lot like Marvin the paranoid android from the BBC TV adaptation of The Hitch Hikers Guide to The Galaxy in the 1980’s. In fact, the BBC Marvin, is about as smart as the 2018 Sophia. They are both controlled, scripted and as smart as a dish washer. Less so.
Hanson Robotics was called out, just last week by Yan LeClun, Facebooks head of AI. In his tweet, basically calling it BS. Out of anyone, most people, he would know, but in many ways, we all, most of us, as those that enjoy and have an interest in tech, should have known and probably did, all along. That Hanson Robotics are selling a dream, that Sophia, even in the words of Sophia’s co-creator, Ben Goertzel, regarding what we see and what is truly real, Goertzel defended it saying that ‘ it encouraged people to believe in AI progress. When pressed further he stated that: Sophia is good publicity for Hanson Robotics.
So there we have it. She’s, it is a side show, not the real event, but a puppet of motors and latex, designed to give the impression of AI advancement, ahead of the curve, ahead of Google, Facebook and others. It’s a facade, a promotion of lies with the intent of promoting the company behind it. In essense, it’s BS dressed up as scientific advancement.
This however is damaging. We like to believe that companies, especially those linked to and developing scientific advancement in those incredible areas of AI, are showing us what they have truly found, discovered and developed. It gives us the hope that we as a race are making real inroads and that these leaps are real, life changing and that we do live in exciting times technologically. In as much as we do live in exciting times and those companies and firms that are developing AI, learning and more, do share, a good percentage of the time, real evidence of their advancements in AI. The trouble is, we as consumers of information, video and more, are lazy, we tend not to discern the source, or it’s intent, and many, simply take it as gospel. By that alone, we as consumers, viewers, passively take that information and many do believe that Hanson Robotics and Sophia, is as smart as we are lead to believe through their promotions and videos. Where in factually, Sophia is simply a marketing ploy, to get us speaking about the company that created it, no more, no less. In that regard, Sophia is a success, Hanson Robotics got what they wanted, but it’s disingenious, it’s factually flawed to the core and outwardly decietful. Leading many to believe that when true AI is presented, by the likes of Google and others, that when those that are singing from the hymn sheet written by Hanson Robotics, that those big firms are wasting billions on AI, if that’s as far as we have come. Not really understanding the fact is, that is where we truly are and that the likes of Hanson Robotics have been selling the dream, claiming it to be reality, throwing in some latex, motors and psuedo science, to promote the company and not AI’s advancement. Can ultimately be damaging to how people perceive advancement in tech and what’s presented, but rather foolishly damaging to Hanson Robotics, who come across as false.
What do you think? Do you feel that Sophia and how it’s presented is harmless, or do you feel Hanson Robotics should be more open and honest about it’s ability and their ability in AI. Let me know.