Yuval Harari: In the Age of AI, Will Humans Lose Their Economic Value?
Yuval Harari, Israeli professor and author of Sapiens: A Brief History of Humankind, (http://www.ynharari.com) gave a speech at Tsinghua University on April 21, 2016. The event was organized and sponsored by the Berggruen Institute as part of its Philosophy and Culture Center. The following is a summary of the ideas presented by Dr. Harari.
If we look to the future, we will see the arrival of a new industrial revolution. Humans have again mastered new, enormous powers, which are much stronger than the steam engine, oil, or electricity; these are biotechnology and computing technology. This type of power is not only used for the production of textiles, food, transportation equipment, or weapons. Rather, the main products for these new technologies will be bodies, brains, and intelligence. For thousands of years, humans have learned how to alter their surrounding environment, domesticate animals, and cultivate plants. Humans have learned how to change the structure of our economy, society, and politics. But there is one thing that hasn’t changed throughout, and that is humans ourselves—the bodies, minds, and intellect of humans today do not have any intrinsic differences between humans of ancient China or even the Stone Age.
The approaching technological revolution will change the bodies of humans. Humans will learn how to design, process, and manufacture bodies, brains, and intellect. Similar to the circumstances in the 19th century industrial revolution, the gains and power of the 21st century technological revolution will not be equally shared. A small minority of countries will lead the way. The gaps brought on by this will be larger than the 19th century and this time, those countries that are left behind may not have an opportunity to catch up.
At the same time, we will also see gaps produced within societies, where the aforementioned tech revolution and especially the biological revolution, are to be the main culprits in creating new forms of inequality. The income gap of the past, the gap between peasants and kings, was economic, legal, and political, but it was not biological. In history, some societies and cultures indeed once made people believe in the privileged classes who possessed superior and nobler “blood lines”. But this was not real, and according to our current understanding of biology, the physiology and psychology between the king and peasants was the same. The differences were simply societal, political, and economic.
But in the 21st century, new technology will grant people unparalleled abilities, giving rise to biological differences between the rich and the poor: the rich elite will be able to design themselves or their descendants, making them advanced “supermen”, psychically and psychologically. Humans will split into different biological classes because of this, where the previous social and economic class systems could transform into a biological class system.
Will people lose their “economic value” in such a new environment? There are some clear trends, such as how increasingly smart computers and robots are forcing people out of their jobs.
Let us use drivers and doctors as examples. A decade ago, there were many experts who claimed that getting around a city was full of uncertainties. The technology for driving in such a situation is too complicated for a computer; a person must do the driving. But today, there are many experts who already expect that in 2025 some cities will have self-driving vehicles. In the future, some cities might go as far to ban humans from driving, as humans will break the law, drive drunk, or drive when they are tired. But perfected artificial intelligence does not have these defects. Artificial intelligence can be connected to the Internet and know the thoughts that humans who drive have no way of knowing. Artificial intelligence can share information with other AI systems and thus avoid car accidents. Yet today, deaths from car accidents are higher than any type of violence. Given these clear advantages, we can expect hundreds of thousands of taxi and bus drivers will lose their jobs as humans people are supplanted by (AI).
Along similar lines, human doctors might possibly disappear. Human doctors can only diagnose off of a smattering of quick questions and tests. They are unable to grasp all of the similar cases across the world, and do not know patients and their family members DNA or medical histories. Human doctors will fatigue, get sick, and get angry. Compared with humans, artificial intelligence can do a much better job. Take the Watson project that IBM is currently developing, which can collect case history from the entire world and can continually update its ginormous database. Watson can get a user and his or her friends’ DNA or medical history. Watson can make a diagnosis through precise data analysis and come up with a treatment. The user can answer Watson’s questions in their home, and can do all kinds of tests at any place and time. Watson can also be with a user at any time and place through smart phone applications or wearable technology, where Watson can observe the users health, test blood pressure and heart rate and give timely medical suggestions. Of course, there are some technical and legal issues with these artificial intelligence doctors that need solving, and thus they won’t take the place of human doctors in the near future. But the advantages of the Watson project are that it takes 10 years to train a doctor and we can only train one doctor at a time, yet for artificial intelligence, as long as we can get past the technical cruxes, the benefits can be replicated infinite times and thus create limitless value.
In the face of these challenges, a common question emerges: artificial intelligence perhaps can be intellectually competent, but what about emotionally? If you get a horrific disease like cancer, would you want a cold machine or a warm-blooded human doctor to administer treatment? On the surface there are two choices, yet truthfully there is a third choice: warm-hearted artificial intelligence. For artificial intelligence, human emotions are purely a biological and chemical process. They can use expressions and sounds to distinguish and respond to the changes in your emotions. Now, some marketing companies already use similar technology to judge a caller’s psychological state. Computing processes can analyze your tone and diction through your voice and then use a huge database to judge your present emotional condition and your type of personality. Then artificial intelligence can combine these two analytic results and match you up with an appropriate service. In this aspect, artificial intelligence apparently can do this better than humans can; they do not have emotions of their own, but can distinguish and respond to yours.
Will humans become useless? Optimists think that even if robots can replace humans in many jobs, there will still be new jobs coming in and humans will transfer over to these new jobs, especially within the service industry. Thus, humans will probably not be totally forced out of the labor market. But there are two basic abilities of humans: physiologic ability and cognitive or psychological ability. From the 19th to the 20th century, robots have on a large scale taken those jobs that require physical capability, such as agriculture, and people have moved to jobs that require more cognitive ability. Yet today, robots have begun to possess this cognitive ability, and are replacing humans in these types of jobs too. Unless humans possess a third type of ability—one that robots do not possess, we cannot take for granted that the 21st century will play out as the 20th century did.