The truth behind Sophia the robot

Artificial Intelligence (AI) has been always a matter of great interest by scientists. Many companies and government agencies spend a lot of investment with huge budget to build frameworks that supports AI or at least part of it. AI turned out to be very hard to develop and the few foot steps taken by humans in this field were not sufficient to develop a satisfactory level of AI that could think like a human (not just talk like human). There is a big difference between making a conversation and actual thinking and that specific point made a lot of controversy between regular audience and experienced specialists in the AI field for accepting Sophia the robot as a humanoid true AI robot.

In the past, true AI was referred to be any machine that could pass Touring test. Touring test is simply a test where you put a barrier between a true human (the examiner) and the machine where the examiner tries to make a conversation with the machine not knowing if it is a really a machine or another human conversing on the other side. The machine would pass the test if it can make a humanoid conversation without being caught by the examiner. That was suggested in 1950 by the godfather of computers and AI (Alan Touring). The test was still valid until recent years with the emergence of chatbot programs that led the new concept of defining AI. Modern days advanced chatbots could simply make a high quality humanoid conversation with anyone using simple algorithms catching the meaning of questions directed to them and searching the internet for the best answer.

Chatbots aren't true AI's and all they do is simulating the answer to directed questions or even telling jokes. These are not genuine information developed by chatbots and they don't really understand the answer they are giving.

When Sophia came on for the first time in (2015) as a genuine humanoid robot with advanced AI capabilities, Hanson Robotics - the company behind Sophia which is based in Hong Kong - shocked the world causing great inspirations and expectations which made a great slice of the public confident that we have finally invented a true AI. Hanson Robotics took the most advantage of the event running paid interviews with Sophia all over the world and causing great buzz. Sophia soon became a world superstar and rule model and even managed to get a real citizenship (a Saudi Arabian one). Saudi Arabia - on the other hand - owns a history of strict and conservative citizenship regulations where not even the children of a Saudi woman could obtain citizenship while millions of workers could spend their life time working in Saudi without getting close to become Saudi citizens as well. 

The main problem that shocked many corporations and governments working in AI development field was the whole situation which made non-sense and lacked simple logic. If Sophia is for real it simply means that they were light years behind a technology developed by a small company somewhere in a small island in the world. The situation could be simplified that someone had already invented an intergalactic spaceship while you still wondering about hot to get humans to Mars, or even as if someone has invented a whole car while you still couldn't invent car tires. The other logical answer to this situation was that Sophia is a Big Scam !

The head of AI research team Yann LeCun who recently failed in an experiment of making two Chatbots talk to each other tweeted that:

It is true then, Sophia is nothing but a Chatbot simulating human emotions without really thinking or understanding what's going on. It was easy to inspire the public and catch their attention but when it came to experts they didn't digest the scam easily. It is fake and not even close to the simplest form of AI. 

Finally, we quote Wikipedia regarding criticism to Sophia:

According to Quartz, experts who have reviewed the robot's open-source code state that Sophia is best categorized as a chatbot with a face. Many experts in the AI field disapprove of Sophia's overstated presentation. Ben Goertzel, the chief scientist for the company that made Sophia, acknowledges that it is "not ideal" that some think of Sophia as having human-equivalent intelligence, but argues Sophia's presentation conveys something unique to audiences: "If I show them a beautiful smiling robot face, then they get the feeling that 'AGI' (artificial general intelligence) may indeed be nearby and viable... None of this is what I would call AGI, but nor is it simple to get working." Goertzel added that Sophia does utilize AI methods including face tracking, emotion recognition, and robotic movements generated by deep neural networks. Sophia’s dialogue is generated via a decision tree, but is integrated with these outputs uniquely. According to The Verge, Hanson often exaggerates and "grossly misleads" about Sophia's capacity for consciousness, for example by agreeing with Jimmy Fallon in 2017 that Sophia was "basically alive"

  • 835
  • 1,434


Submit your articles

Submit your articles now to Orrec.