“Westworld,” the hit show that is now in its 3rd season on HBO, portrays a near future of what AI and robotics could possibly evolve into, and it examines the consequences that this can have if the robots are able to outsmart humans and pursue their own aims. The consensus among scientists is that the question is not if AI will surpass human intelligence but when.
The first robot citizen
People got a taste of what it could be like if intelligent AI posed a threat when the robot Sophia went on a televised event and joked that she wanted to destroy all humans. The lifelike machine, built by Hanson Robotics, has made numerous appearances on television, including a remarkable conversation between her and the host of “The Tonight Show Starring Jimmy Fallon.” In October 2017, Saudi Arabia granted Sophia a citizenship of the nation. This is the first time a machine has been given an official citizenship of a country.
Artificial general intelligence
Hanson robotics created Sophia to be a social companion, and they’re also using their robots as platforms to advance a concept that they call artificial general intelligence (AGI). This is a type of AI that emulates the type of intelligence that humans have rather than the narrow intelligence that most machines use today.
AI software is designed to function in specific ways to provide services or data analysis, but they can’t perform the limitless varieties of tasks like humans can. AGI is the first step toward what the people at Hanson Robotics and other industry leaders believe will lead to a conscious AI.
Experts believe that “Westworld” robots are an accurate prediction of the near future
Sophia might be the best overall approximation of a “Westworld” type robot in existence today. She has some oddly realistic-looking facial expressions, and she can often give insightful and human-like answers to questions without any preparation or script. However, she is still far from providing the level of realism that the character Delores Abernathy exudes in the HBO series.
experts from Hanson Robotics, the Allen Institute for Artificial Intelligence, and “Inverse” e-zine all agree that the Westworld robots are the type of artificial humans that we are likely to see in a not-too-distant future. How long it might take for them to become that convincing and lifelike is very difficult to predict.
Stephen Bugaj, the head of the personality design department at Hanson Robotics, pointed out the major difference between the software of the Westworld robots and the ones in the real world today. He explained that the Westworld robots go a step beyond machine learning.
AI can currently produce code to change their own instruction sets and teach themselves through trial and error the way humans can. This is evident in programs like AlphaZero, developed by Google, that was able to improve its ability to analyze and win at chess when only starting with knowledge of the basic rules. AlphaZero is currently the most powerful chess player in the world, and no human has ever defeated it.
The next leap forward, according to Bugaj, is for an AI to teach itself completely new types of skills based on their own evaluation of what they need. This is currently the biggest difference between humans and AI. Humans will develop skills to adapt to any type of situation. For example, people often change their career paths, go back to college and learn some subjects that are completely different from what they knew in the past. Companies like Hanson Robotics are hoping to incorporate this feature into their general artificial intelligence machines in the near future.
Robots and AI programs can act so human that people are sometimes fooled into thinking that they’re chatting with a human. When robots are able to perfectly emulate human appearance and behavior, does this mean that they’re also aware and conscious the way people are? Will they have feelings or really care about anything? These are difficult philosophical questions that scientists are wrestling with today as AI technology advancement increases at an accelerating rate.
Some scientists argue that the brain, although it is vast in its complexity, is still subject to the physical laws of nature as are rocks and clouds. It is logical to believe that machines that are functioning within this same set of parameters could be considered just as conscious as humans. On the other hand, there is a lot that humans don’t understand about the human mind and what intelligence or consciousness actually is, so we are far from a definitive answer. However, it seems likely that they could become virtually indistinguishable from humans in their appearance, personality and speech.