“The Media Overstate The Capabilities Of Artificial Intelligence”

“Recently, executives often ask me, what is now the technology of artificial intelligence. They want to understand how to use such technologies in the work of their organization and how they will affect their business,” says Andrew ng. Media, says Eun, often greatly overstate the capabilities of modern artificial intelligence technologies. Many of the media tell readers that development can almost take over the world in the coming years.

“In fact, AI really is used in many areas. Web search, advertising, e-Commerce, Finance and so forth.”. “As one of the founders of Google Brain and former head of the research laboratory in the field of artificial intelligence at Stanford and the current head of the AI-direction in Baidu, I worked with many teams involved in the development of such technologies and to participate in the construction of a large number of products based on them. And thats what I can tell.

Artificial intelligence will change many industries. But there is no magic here.”. Eun offers to understand what modern technology is capable of artificial intelligence — in his words, it will help to understand how they will affect business in the future. As noted by the entrepreneur, the possibility of artificial intelligence is still limited. And almost all of the existing technology based on the ability of the AI to quickly process input information and give A simple answer B.

Examples: As noted by Eun, the ability to answer according to the information And will transform many industries. “For such technologies is the technical term. Learning with a teacher. And they are far from intelligent robots, the appearance of which we read science fiction.

The human intellect does much more. Such systems are evolving rapidly, but they are still far from what is shown in movies.”. Scientists, says the entrepreneur, are developing other systems of artificial intelligence. “Some of these technologies are well show yourself in some cases. Its possible that the developers will make a breakthrough, but it is not clear how to go to this”.

As noted by Eun, todays systems have a weak point. They need to work a huge amount of data. “We need to demonstrate technologies many specific examples and solutions that have learned to get the result. For example, to train the system to recognize faces in images, it is required to drive hundreds of thousands of images, together with the prepared answers. To create a speech recognition system needs a knowledge base of not less than 10 thousand hours of audioredo together with ready transcripts”.

“But what can the systems that according And get an answer In. There is one rule that helps to understand the breadth of application technologies. If a person can solve a problem not more than a second thought, most likely, in the near future it will learn to solve artificial intelligence technology”. According to entrepreneur, there are many important tasks that now faced people. For example, to identify on the videos suspiciously behaving people, finding and removing offensive entries in the network, to make a decision about where to direct the car.

The human brain spends on the decision no more than a second. “These tasks can be automated”. Work with artificial intelligence requires careful selection of source data and desired result. But such technology can truly revolutionize many fields. Eun believes that once leaders realize what can and can not do artificial intelligence, they can understand how to apply these technologies in business.

“In this area just take and copy someone elses solution — often ineffective step. You need to understand what value the technology brings to a particular company. But the community around the technology of artificial intelligence is surprisingly open, it is common to exchange ideas and even developments.”. Among the most of the limited resources of industry data and talents. “In current conditions, one team can replicate the results for another 1-2 years.

But to copy someone elses data. That is data, not software are the obstacle for many companies. And talented engineers are able to not just copy someone elses software, but also to adapt it to the company and to apply it in the best way. So for talents has started a real war.”. According to the founder of Coursera, the biggest harm in the near future may bring artificial intelligence to the man — reduction in the number of jobs.

“The amount of work that can be replaced by technology is growing faster than ever. And our task as leaders is to make sure that in the new conditions everyone will find a use”.

Leave a Reply