Understanding artificial intelligence from a social perspective
Wu Jun/Text
I am very honored to write this introductory article for "The Era of Artificial Intelligence and the Future of Humanity". Using the author's language, I can decipher what's most important about the book and how it affects us. This book is very useful for us to understand artificial intelligence from a social perspective.
In the thousands of years of human history, geniuses have created countless amazing ideas. They laid the foundation of human philosophy and contributed to the progress of science. Every historical period has left behind its own historical achievements, from the period of ancient civilizations to the period of the Enlightenment. Today, even children are familiar with the names of great thinkers such as Confucius, Buddha, Jesus, Aristotle, Khwarezm, Shakespeare, Newton, Beethoven and Einstein.
However, today, there is a new group of subjects of invention and creation, and they are not human beings! In many ways, artificial intelligence has surpassed humans and exceeded people's wildest imagination, such as in various games. Thousands of years of human understanding of Go have been transformed by AlphaG. The depth of this level of telling that it is wrong is obviously far out of reach compared to the depth of human understanding of Go represented by the latter. Of course, what is more realistic is to use deep neural networks to produce textured images, write smooth and beautiful articles, and design the molecular structure of drugs. This algorithm can be used to autonomously drive and control fighter jets. It can not only replace the driver of commercial vehicles, but also enable the operation of military aircraft. All in all, we have never felt so excited and yet so threatened, not just because they threaten our jobs, but because we fear that we will be manipulated by completely unknown forces.
In addition, current artificial intelligence algorithms will continue to amplify errors in training, because the training of various algorithms requires millions of iterations. For example, in the medical field, although artificial intelligence technology has been used to study the different results of the same treatment (or drug) produced by different people's genetic genes, this research on personalized medicine based on artificial intelligence has led to over-matching of treatment plans. Phenomenon, it works very well for some people, but completely ineffective for others. This is because the training data for artificial intelligence comes from
Some people, and it doesn't know how to adjust the results for others.
Although these "half-baked" artificial intelligences are not perfect, they are already starting to change the world and are turning us into different people than we were before. For example, Google’s search engine or Facebook’s social network, which consumes billions of dollars every day, use artificial intelligence to censor various content, filter results, and provide us with the society they want us to see, rather than the complete society. Whether this is a good thing or a bad thing is still hard to say. Regardless, we have changed as a result of their influence.
Today, people’s attitude toward artificial intelligence is similar to that toward nuclear technology nearly a century ago, with both fondness and ambivalence. On the one hand, nuclear technology gives people hope in many fields. It can not only solve human energy problems, but also be used in medical and other fields. People may also use nuclear technology to carry out destruction and nuclear blackmail, which directly threatens the survival of mankind. Just like how we viewed nuclear technology in the past, we are now unable to judge the pros and cons of artificial intelligence technology. The designers and users of technology determine the boundaries between good and evil, but it is difficult for us to use unified standards to measure their moral behavior. All we know is that scientific progress won’t stop just because we don’t like it, and trying to slow down the discovery of AI technology is a fool’s errand. What we can do is that as the main body of society and as leaders, we must figure out what fields artificial intelligence technology can be used in and where it should be restricted.
Today, even this level of "half-baked" artificial intelligence is already quite powerful. In the future, it will become an almost omnipotent existence. All we need to do is to make good use of it. For example, this technology should be used for drug research and discovery of new drugs, but its use in cyber attacks and wars should be strictly limited. We must ensure that the development of artificial intelligence is consistent with human ideals and well-being. For this reason, from now on, we must effectively manage artificial intelligence technology. Artificial intelligence technology has become popular all over the world, and the learning threshold is lower than that of nuclear technology. We are no longer able to formulate a norm similar to the Treaty on the Non-Proliferation of Nuclear Weapons to prevent someone from using artificial intelligence to do evil, but it is still very important to set a standard for the reasonable use of artificial intelligence around the world.
We need to ensure we educate the next generation to use this technology to advance human welfare and serve the public good. While some may scoff at our seemingly naive notion, it is crucial that we take the first step.
Today, mankind stands on the crest of the wave of technological revolution. Our generation is fortunate to usher in the intelligent era, but we must also assume the responsibility of guiding the healthy development of intelligent technology.
(The author is a computer scientist and Silicon Valley investor. This article is an introduction to the book "The Age of Artificial Intelligence and the Future of Humanity". The authors are Henry Kissinger, Eric Schmidt, and Daniel Huttenlo Hull)
The above is the detailed content of Understand artificial intelligence from a social perspective. For more information, please follow other related articles on the PHP Chinese website!