Artificial intelligence (AI) is becoming the next industrial revolution, and it will quickly reshape the business, the global economy, and the way people work and interact. In an interview with foreign media, Wu Enda, Baidu’s chief scientist and co-founder of education startup Coursera, and Neil Jacobstein, head of artificial intelligence and robotics at Singularity University in Silicon Valley, discussed labor. Opportunities and challenges facing intelligence. Wu Enda (born in 1976, English name: Andrew Ng), Chinese American, is an associate professor in the Department of Computer Science and Electronic Engineering at Stanford University, and director of the Artificial Intelligence Laboratory. Wu Enda is one of the most authoritative scholars in the field of artificial intelligence and machine learning. Wu Enda is also the co-founder of the online education platform Coursera. On May 16, 2014, Wu Enda joined Baidu as the chief scientist of Baidu, responsible for the leadership of Baidu Research, especially the Baidu Brain program. Jacques Bustan is the head of the singularity university artificial intelligence project, the US Department of Defense and NASA consultants. He has worked on Stanford's augmented decision system and served as chairman of the Artificial Intelligence Conference on Innovative Applications, with a background in environmental science and molecular biology. Singularity University (SU) is located in the heart of Silicon Valley, California. In the NASA Ames Research Center, the era of computers is better than the human brain. Google (microblogging) and NASA (NASA) ) Collaboration is a school that trains future scientists. It aims to solve "a major challenge facing humanity." His research interests include synthetic biology, nanotechnology, and artificial intelligence. He began enrolling students in June 2009. The institution is led by renowned futurist Ray Kurzweil, who includes Neil Jacobstein, who specializes in artificial intelligence projects, and Brad Templeton, who is also a network unmanned project consultant. Raymond McCauley specializing in biotechnology and information projects. Wu Enda: For large companies like Baidu, AI offers two big opportunities. The first is our core business. Web search, advertising - nowadays it is driven by AI. Let's say we run a very large meal delivery service. When you order a meal, we use AI to predict how long it will take to deliver the meal to you. This includes deciding which motorcyclist to deliver the meal, so that when he arrives at the restaurant, the food is just ready to be delivered to your hand, and they can be as hot as possible. In addition to enhancing our core business, AI is creating many new opportunities. Just like about 100 years, electrification has changed almost every major industry, and I think it is now in the stage where AI will change almost every major industry. Therefore, part of my work in Baidu is to systematically explore new vertical areas. We have formed an autonomous driving department. We also have a computer that can talk, similar to Amazon's Alexa and Google Home. We are systematically entering new industries because we feel that in these industries, an AI team can be formed to create and capture value. Wu Enda: A few years ago, we started to invest heavily in speech recognition because we felt that its accuracy would reach a very high level, making it a normal state for users to use it. A few years ago, we achieved 95% accuracy, and the difference between it and 99% accuracy is not just an incremental improvement. This difference is a bit like a double-fired day. When the accuracy is 95%, you will hardly use it, just like a few years ago. Once the accuracy is improved, you will always use it, naturally. Become a habit. In Baidu, we have crossed the inflection point of the user's adoption curve. In the past year, the annual growth rate of daily active users using voice recognition in Baidu's various services has reached 100%, and we expect to continue to grow. After a few years, using speech recognition will become the norm for everyone. Soon you will forget what it was like when you couldn't talk to a computer before. Jacabstein: In 2017, we saw that the AI ​​developed by a team at Northwestern University performed better than the average US in the Raven Progressive Matrices Test (Intelligence Test for Visual and Analogical Reasoning). People are better. We also saw a team at Imperial College London developing an AI that diagnoses pulmonary hypertension more accurately than a cardiologist. The accuracy of a cardiologist is about 60%, and this AI is 80%. In January of this year, Tuomas Sandholm and Noam Brown from Carnegie Mellon University developed a poker AI called Libratus, which won four world championships. Poker hand. They played 120,000 hands of poker and Libratus eventually won $1.77 million in poker chips. This is a big deal, because in the poker game, you are faced with incomplete information, sometimes you need to bluff, the opponent can also mislead you with the wrong information, and the AI ​​shows the skills to deal with this kind of situation. This is a very important skill that is useful in the areas of negotiation, strategy development, and even policy analysis. Wu Enda: Modern industry has experienced winter, winter, winter, and then eternal spring. I think that we are now in the "eternal spring" stage of AI, because today's AI is creating great value for companies like Baidu and Google, unlike the previous wave of speculation. This opens up a very clear source of income that allows AI technology to continue to invest and improve. Wu Enda: At the moment, AI technology is a magical thing, right? It is useful for many different things. But the reality is that AI technology needs a lot of customization based on your business background. So I recommend that the corporate leadership hire a senior AI executive, such as a chief AI officer or vice president, to help them sort things out. Recruiting AI talent is very difficult. If you want to be consistent in the AI ​​team's recruitment, promotion and management standards, then the best practice is to set up a centralized AI department. Then let the team work across departments to figure out how to incorporate these technologies into the company's business. Jakabstein: I have a different view of this matter. I believe that small, interdisciplinary, and high-level support teams will have a lot of power. Matching the speed of AI technology with the flexibility of the AI ​​team is very important. Setting up a centralized AI executive at the top of the company, everyone must ask the supervisor about the problem, and it is unlikely to be as fast and productive as a distributed architecture. Wu Enda: Although the situation may change in the future, there is a rule of thumb today: almost anyone who can do it with less than a second of mental thinking, we can use it now (or in the near future) AI automates it. This is far from being able to get all the work done. But there is a lot of work that can be done by connecting a number of "one-second tasks" together. For example, a guard who is responsible for viewing surveillance video, his work is very complicated. But this work may be broken down into many smaller tasks, all involving one second of cognitive thinking. So figure out where to use AI. This skill and art is to identify business opportunities to a large extent. Your system is complex, but it also contains many such "one-second tasks", you may be able to put them Connected in series to automate. Jacabstein: I think that the speed with which machine learning replaces everyday work can be shocking. Moderator: How fast? Jakabstein: We are talking about the changes that will take place in the next 10 to 15 years. This is very important. Therefore, we need to invest heavily in free education and explore ways to provide basic income to those who have lost their jobs. Wu Enda: Just as AI will destroy work, it will create new jobs that we can't imagine. The challenge is that the skills don't match. Jakabstein: We are talking about a very significant shift in the next 10 to 15 years. Therefore, we need to invest heavily in free education, while exploring ways to provide basic income for those who have lost their jobs. Wu Enda: AI will steal people's work, and it will also create new jobs that we can't imagine. The problem is that people's existing skills don't match new jobs. Jacabstein: The good news is that AI and robots can create a lot of new wealth. Our responsibility is not only to ensure that the company is successful, but also to make people who have lost their jobs still enjoy a reasonable quality of life. Therefore, we need to lower the price of education so that people can afford it because there will be new jobs. But the real question is, “What is the ratio of work done to new jobs?†I think that at least in the short term, this ratio may be bad for people. Marine Diesel Generator,Marine Generator Set,Diesel Marine Generator,Boat Genset Jiangsu Vantek Power Machinery Co., Ltd , https://www.vantekgenerator.com
