Artificial intelligence, the process through which machines can learn human-performed activities, poses one of the main challenges of the 21st century. Experts agree that the new technological revolution will impose complex problems for governments and people alike and that the response to it should involve both international measures and individual solutions.
In his new book, ” 21 Lessons for the 21st Century“, Israeli historian Yuval Noah Harari warns that what he calls the fourth industrial revolution might make it harder for governments and workers to adjust to. Unexpected countries may become leaders in AI use and application, new alliances might form while some traditionally strong economies might suffer irreparable damage. As AI shapes a new relationship with the world around us, it will also force us to embrace more creative jobs that can’t be easily automated, and learn new professions almost every decade. The future, Harari says, is for the creative and the resilient.
The interview has been edited for length and clarity.
Can we already predict the winners and losers in the global AI race, or can countries still catch up?
It’s not completely predictable because it depends not just on pure technological and economic factors, but on cultural and political factors. We are already in a place where we can see some winners and losers. Some of the Western European countries that led the industrial revolution are quite behind in the development of artificial intelligence and machine learning.
What we are really likely to see is maybe a repeat of the 19th century industrial revolution when a few countries industrialized first, most countries were far behind and the industrial powers, even if they were quite small countries, conquered and dominated and exploited everybody else. Relatively small countries like France or Italy — even Belgium — conquered huge empires, whereas very large countries like India or China were dominated by the industrial powers. The same thing might happen again in the 21st century with the automation revolution on an even more extreme scale.
Can AI make people irrelevant?
With the industrial revolution, the big danger was that some countries and some classes will exploit others. In the 21st century, the biggest danger is that many people and even many countries might become irrelevant in economic terms and therefore also politically very, very weak. If a country has natural resources like oil or iron, it’s likely to still be valuable, but the working power of the people might become much less relevant as machines and computers learn to outperform humans in more and more tasks. Some humans will still be necessary and will become even more powerful than today, but the danger is that a certain percentage of the population, even in developed countries, and maybe the entire population of some countries, will lose their economic power and importance and therefore also the political power.
What measures can countries take to counteract the negative effects of this new revolution?
If a country that relies almost exclusively on manual labor, let’s say the textile industry, lost almost all of its economic power and people are unemployed and the economy crashing, the only thing that can really help is an international safety net. You have a lot of people now talking about universal basic income, but most of them actually mean national basic income. They have in mind a situation (that could happen in) the United States where you tax the big high-tech corporations in Silicon Valley that make a lot of money from the automation revolution. The U.S. government uses these taxes to support, say, unemployed taxi drivers in New York or unemployed coal miners in Pennsylvania.
But the biggest problems will be in countries like Guatemala or Bangladesh. Will we see a situation in which taxes on the high-tech companies in California or in eastern China (are used to) support unemployed people in Bangladesh? This is a very unlikely scenario, but if people talk about universal basic income, they should really mean universal, and not national.
What would an international response to automation look like?
Many countries don’t have the resources and the infrastructure to develop by themselves an IT sector, but they can work together to compete with the U.S. or China. If it’s unlikely that countries like Uruguay or Chile or Colombia build its own Silicon Valley; maybe they can do a regional cooperation to develop a South American Silicon Valley.
Another option is at the very least to cooperate in order to increase the power of negotiation. A lot of the development of AI is based on the accumulation of enormous amount of data. This data is being accumulated from throughout the world, from countries like India, Brazil or Mexico, and is centralized in places like California or China. You can say that to some extent what we see today in the world is a kind of data colonization. Just as in the 19th century (when) raw material from South America, Africa, Asia, was used to fuel the industrial revolution in Europe and the countries that provided the raw material gained almost nothing in return, a similar thing may be happening now with data. A single small country can’t do much about it, but if several countries say, OK, we can’t compete with Google or Tencent, but we are giving you our data and we want something back, this may work.
You often mention AI can lead to a digital dictatorship. What is this?
(This is a) situation in which almost all the data about people, their habits, their personality is centralized in the hands of the government or some corporations. That government or corporation is able to control and manipulate the lives of people more than any dictatorial regime we’ve seen before in history. This may sound like science fiction, but it’s present data. An emotion like anger manifests itself in all kinds of physiological and biochemical processes in the body, like a higher heart rate or blood pressure and certain brain activities, certain muscle (activities). (So you can have) this Big Brother that has access into your brain and into your mental states every moment of the day. It’s not like we are going there inevitably, but if we’re not careful about these new technologies, this could be a plausible result.
You also mention that our education system needs to fundamentally change. How will that happen?
You need a brilliant education system that doesn’t stop at 18 or 24. You need an education system throughout your life. We need to think about education and the re-skilling and retraining of adults even when you’re 40, 50 or 60 because of the rapid changes in the job market and in society as a whole. It also means that the emphasis in education needs to shift from giving students information and teaching students skills. The emphasis should be on emotional intelligence, on mental resilience, on learning how to learn, on building a kind of flexible personality that enables you to keep change and keep learning throughout your life.
What type of jobs could people be asked to master in the future?
Nobody really knows. It is likely that there will be many new jobs. The real question is whether people will be able to retrain and reinvent themselves to fill the new jobs, especially because the new jobs are likely to demand a high level of creativity and flexibility. Any job that is very routinized will be very easy to automate. Then the question is whether people will (be able to) retrain and fulfill the new jobs and do it not just once but several times during their lifetime. Because it won’t happen just once, but maybe every 10 years, as AI continues to develop. The new job (they get) will also be automated or changed within 10 or 15 years, so maybe you’ll have to reinvent yourself four, five, six times during your lifetime, especially with lifespan increasing and people retiring at an older age. The big question is really psychological.
Is creativity and mental resilience something you can teach?
I think that yes, but it’s much more difficult than teaching dates in history or equations in physics. The big problem is that we don’t have the teachers. You can train people who guide people to develop their emotional intelligence and flexibility, but you need a very different kind of teacher than many of the teachers we have today. We are in a vicious circle. The teachers we have today are the product of the education system of yesterday and many of them don’t have the necessary skills for the kind of education in the 21st century that we need. So maybe we need to begin by retraining the teachers, but who will retrain them? There is no easy solution to this question. It’s really going to be quite a difficult transition.
More from U.S. News
AI Might Force You to Learn a New Profession Every 10 Years originally appeared on usnews.com