Thoughts on the application, development and influence of artificial intelligence

Time:2020-11-18

Introduction:The artificial intelligence research center of Cambridge University analyzes and discusses the new functions and risks of artificial intelligence, and corrects some people’s doubts and prejudices about artificial intelligence.

The artificial intelligence research center of Cambridge University analyzes and discusses the new functions and risks of artificial intelligence, and corrects some people’s doubts and prejudices about artificial intelligence.

The research center brings together experts from different fields to study and predict the possibilities and challenges brought by the rapid development of artificial intelligence, and provide a more measurable and useful perspective for artificial intelligence. Dr. se á n ÓÓ h É igertaigh of the center for human survival risk research (CSER) of Cambridge University, who participated in the formulation of the new proposal, pointed out: “artificial intelligence occupies an important position on the agenda of our research center, which is partly due to the significant progress of artificial intelligence technology in recent years. This shows that its research has had a great impact, and progress is very rapid. On the other hand, because AI has a lot of problems to deal with, just focusing on disasters and risks limits the scope of our research in this field. “

The idea of the research center is to bring together AI experts from relevant disciplines to consider not only long-term risks, opportunities and challenges, but also the long-term, short-term and medium-term impacts of AI.

Thoughts on the application, development and influence of artificial intelligence


Different methods of artificial intelligence

Although artificial intelligence often makes headlines in industry media or the theme of some science fiction movies, the research work of CSER at Cambridge University provides different perspectives and perspectives.

While the risks associated with AI are chilling, the current impact is relatively limited. Dr Ó h É igertaigh explained: “frankly, the adverse effects of AI systems that people see are limited, because they are mostly good practices for performing certain tasks, such as traffic navigation, playing chess or running search engines. At present, we may be able to solve the problem of cognitive ability in artificial intelligence. “

This prudent approach not only makes the centre aware of the serious problems that artificial intelligence technology can bring, but also provides opportunities to take advantage of this new technology.

“In this world, only living things can learn, adapt, think and do more,” said Dr. h Ø igertaigh. But defending the idea that intelligence can only happen in biology proves the argument that at some point, we will have enough resources to recreate it. “

Although his main research area is computational biology, he has been engaged in interdisciplinary projects for many years. In this way, he adopts a multidisciplinary way of thinking, dealing with the problems of artificial intelligence from various angles.

Dr Ó h É igartaigh pointed out that “the answer to these major problems is not only in the field of computer science or computational biology, but also to think about these long-term and wide-ranging problems, which requires the use of expertise in political, economic, legal, sociological and even philosophical fields.”


Application of artificial intelligence

In fact, the need for different perspectives of AI is not only a part of generating original ideas and ideas, but also part of the answer to how AI uses different expertise to solve problems.

“Most of the challenges facing scientists are that they have to analyze large amounts of data from a variety of sources and understand the extremely complex systems that are interconnected,” said Dr. h É igartaigh. It’s a very difficult thing even for multiple teams working on collaborative research.

The system we are currently developing can handle big data. For example, help analyze millions of genomes to find the origin of cancer, analyze many aspects of climate change, or try to make solar energy, energy grids or smart homes more efficient. If we find out how to apply artificial intelligence to the problems we face, we can solve these problems. We will also contribute to the progress of mankind. “

The social, political and cultural aspects of accelerating technological change are also scientific issues. Taking the short-term problem of self driving cars in the future, which leads to the unemployment of taxis or long-distance bus drivers, for example, artificial intelligence takes the place of human resources and encourages them to seek other jobs. This is why different fields should discuss these issues.

Dr. Ó h é igertaigh pointed out that although there are some risks to be addressed (for example, artificial intelligence will soon make the development of multifunctional UAVs a reality), artificial intelligence is not equal to human intelligence.

Unlike artificial intelligence, which is limited in many technologies, many failure predictions for general artificial intelligence have appeared in the past. “Some may argue that the current enthusiasm for AI development is wrong,” said Dr. h é igertaigh. And we’re going to see more exciting investments in artificial intelligence. Although it is only 50% likely that this will happen in this century, someone should be thinking about it and studying it. “

This also shows another important point: even if the AI technology fails as a whole, the technological progress made in this field is still very important, and the social, cultural and political impacts of these developments and advances in AI technology need to be considered, discussed and considered.


Different types of intelligence

Another issue involved in the discussion and discussion of artificial intelligence is that people deal with this problem in a human centered way. However, different types of intelligence exist in the world.

Dr Ó h É igertaigh suggests a way to put both human beings and the earth at the center, from human intelligence to carnivorous intelligence, rather than limited to human centered intelligence.

Dr Ó h É igertaigh said that the first project they defined in the initial phase was “smart type” and had already started meetings on this project. Experts included Murray Shanahan, a professor of Neurology at Imperial College, who specializes in chimpanzee intelligence, mathematical logic and machine learning. All the experts attending the conference are committed to finding relatively new ideas for different types of intelligence.


How to develop artificial intelligence

Another question, says Dr. h é igertaigh, is how this AI will develop. Evolutionary biology is developed through repeated experiments. Some organisms with high error rate develop faster than others with low error tolerance. “When designing algorithms and AI, experts can choose the way they want,” he said. We also offer artificial intelligence learning courses, which we call evolutionary algorithms, to some extent using trial and error. There are also reasons why we want to be open to what happens, and there are reasons why we don’t want change, because we may not get anything important in the end, or it may have unexpected consequences. “

At this point, many different evolutionary factors play a role. The revolution in the field of science has brought more talents and allocated more resources to the field of artificial intelligence, thus promoting the development of artificial intelligence at an explosive speed. “An example of this is the great success of deep learning at an early stage,” he said. This has allowed more resources to be used, and many successful organizations have used this approach. “


Future development of artificial intelligence

Similarly, it can be said that there are breakthroughs in the concept of artificial intelligence, but it is impossible to predict how long it will take or to what extent these breakthroughs will promote the development of this field. Dr Ó h É igertaigh commented: “things we can’t predict bring great uncertainty, so it is certain that general artificial intelligence will be implemented in 2070, which is absurd. But sooner or later, there will be a revolutionary breakthrough. In this case, people need to be encouraged to think creatively and creatively about things with social influence. “

He said that CSER will also serve as a communication center to actively communicate with academics and industry, and hold seminars and industry conferences.

“We aim to create a community that is committed to encouraging future opinion and research leaders to address issues that affect all of us in the long term,” he said. I believe that more and more young people will become future industry and political leaders in this field and will play an important role. “

Source:https://ai.51cto.com/art/202009/625306.htm
The article is transferred from 51CTO, and all viewpoints in this paper have nothing to do with the circle of machine intelligence technology