On the afternoon of April 2nd, the audience of wuzhou Hall, which can accommodate nearly 1,000 people, was reduced by 1/4. In the morning, the audience was full, and some people still wanted to squeeze in, because the three major BAT leaders were all present. It seems that artificial intelligence is not engaged in science and technology without saying a word. However, artificial intelligence needs to calm down and face its development stage and possible problems.

Giiso Information, founded in 2013, is a leading technology provider in the field of “artificial intelligence + information” in China, with top technologies in big data mining, intelligent semantics, knowledge mapping and other fields. At the same time, its research and development products include information robot, editing robot, writing robot and other artificial intelligence products! With its strong technical strength, the company has received angel round investment at the beginning of its establishment, and received pre-A round investment of $5 million from GSR Venture Capital in August 2015.

Zhu Min, president of the National Institute of Finance at Tsinghua University and an economist standing on the sidelines, asked a series of 10 questions to make tech leaders reflect on the ethical challenges posed by ARTIFICIAL intelligence. Shen Xiangyang, executive vice president of Microsoft, pointed out from a professional point of view what aspects of artificial intelligence still need to be improved and breakthrough; Ma Huateng and Li Yanhong talked about their companies’ consideration and outlook on ARTIFICIAL intelligence.

Although ARTIFICIAL intelligence will overturn every business model and may bring various problems, it will also solve many problems as a result. With the participation and cooperation of enterprises, schools and governments, people will come up with solutions to artificial intelligence.

Although after the closed-door meeting, Wu Ying said that there would be a shocking word about artificial intelligence released, perhaps the closed-door meeting leaders research out of the current ai summary. But by the end of the meeting, there were no shock words. Instead, It was Ma’s speech that caused bursts of applause. According to Wu Ying, this time please Ma Yun came over and did not limit the topic and direction, let it want to say what to say. Ma Yun is undisciplined ridicule opponents, preaching Ali’s ideas, explaining the outside world’s misunderstanding of the Internet economy. For the theme of this year’s conference artificial intelligence, Ma said there is still a lot of space.

But at least ai has started.

Is AI a bubble or a false fire?

Someone in the bathroom of the IT Summit sighed that the topic of this year’s college entrance examination must be artificial intelligence. At this year’s conference all the guests, all the audience, all the media were talking about ARTIFICIAL intelligence. At last year’s IT Leaders’ Summit, everyone just mentioned AlphaGo. In just one year, not only mentioned ARTIFICIAL intelligence, every enterprise launched an ARTIFICIAL intelligence strategy.

Although artificial intelligence is not a new word, Shen Xiangyang studied artificial intelligence when he was a graduate student, but artificial intelligence was relatively cold at that time, and he might not even find a job after graduation.

But thanks to the advent of the Internet, Internet + iot provides more data; Moore’s Law led to increased computing power and new methods of calculation; The last five or six years have seen a sudden breakthrough in deep learning. AlphaGo’s victory over humans has made artificial intelligence widely known.

The phenomenon that shen xiangyang mentioned that he could not find a job by studying ARTIFICIAL intelligence is not happening now, but enterprises can get investment as long as they bear the word artificial intelligence.

In March, Intel announced a $15 billion acquisition of Israeli information technology company Mobileye, which some said was overpaid. But as Guo said, Intel’s purchase of Mobileye gives it an important piece of self-driving technology.

The question about ARTIFICIAL intelligence is still, is there a bubble? Wu Ying said artificial intelligence is not too frothy, but not frothy enough, and more bubbles are needed.

But I have to say, in the era of “artificial intelligence”, it is a bit too much.

Ma directly joked that when people around him said that AlphaGo defeated human beings, very scary appearance. So TM What? ! Applause. Ma Yun thinks that if the machine can do a good job, it is nothing. The machine defeated the man in Go, but it insulted him. What the machine needs to do is the thing that the human can’t do, which is the ability.

In this regard, Robin Li also expressed very calmly that the current artificial intelligence, machine learning, deep learning algorithm is still in the very preliminary stage, there is still a lot of room for improvement.

“I don’t think we will be able to achieve strong ARTIFICIAL intelligence. Not only will we never understand how the human brain works, but even if we try to simulate the human brain using computer methods, we will never be able to achieve the level of the human brain. We will never be able to do this.” Li yanhong said.

Ai will disrupt every business model

Artificial intelligence is still in its infancy, but that doesn’t stop companies from getting their hands on it.

Li revealed that Qi Lu, now baidu’s president and chief operating officer, talked to Li about deep learning in 2012, when He was still at Microsoft, and that the technology was close to being put into practice. Therefore, in January 2013, Baidu announced the establishment of deep Learning Research Institute, which is the first research institute named after deep learning in the global industry.

In May 2014, Baidu announced the establishment of Baidu Silicon Valley Laboratory, and invited The youngest of the world’s three major ARTIFICIAL intelligence experts Andrew Ng (resigned) to take full charge of it; In the same year, Zhang Yaqin, former senior vice president of Microsoft worldwide and chairman of Microsoft Asia Pacific R&D Group, joined Baidu as president, responsible for baidu’s emerging business; This year, Baidu also brought in Qilai Lu, a Microsoft executive vice president, as president and chief operating officer of the Baidu Group. Baidu is using AI internally first.

Li revealed that baidu Tower uses face recognition technology while other companies are still using time cards to punch in.

Compared with Baidu, Ma said Tencent is still lagging behind in AI, and only started setting up its Tencent AI lab last year.

Ma pays more attention to ai scenarios, big data and computing power. However, he also said that the lack of artificial intelligence is the most talent, through more than a year, Tencent hired a lot of people, including people from Microsoft. Tencent put the lab in Seattle because the Microsoft people didn’t want to leave Seattle. Tencent’s AI department now has more than 30 scientists, more than 90 percent of whom are PHDS or above and come from Harvard, MIT, Columbia and other universities. Tencent AI is mainly based on basic research in four vertical fields: focusing on machine learning, natural language processing, speech recognition and computer vision.

Zhang Shousheng, chair professor of physics at Stanford University and member of the Us National Academy of Sciences, observed the progress of AI in China and the US. He said the US is still very ahead in AI. But China has a big advantage in algorithms, because algorithms require math, and Chinese math education is very good, and many of the world’s leading AI players are Chinese. As a result, China’s AI status will slowly converge with that of the rest of the world, where currently we see China and the United States competing.

Enterprises have seen the promise of AI, shen xiangyang said AI is now at a very good stage, every business application will be overturned by AI.

When artificial intelligence meets ethics

Ma yun said that he has seen the industry of artificial intelligence for a long time, but heard that many scholars and experts and brain neuroscience experts have also entered the field of intelligence, a little puzzled, human own knowledge of their own brain is less than 3%, human let machine to learn the 3%, a little puzzled.

In fact, shen Xiangyang, an expert in artificial intelligence, also said that artificial intelligence has two branches: one is computer perception, one is computer cognition. Computer perception is going to grow very fast in the next 5-10 years. But cognitive aspects, such as natural language, acquisition of knowledge and thinking and feeling about situations, are not very clear.

“In fact, if you look at the whole development of artificial intelligence, the biggest problem today is that we don’t know much about the human brain. Brain science is still a very preliminary science, because every time we talk about science, we need to have data, we need to be able to do experiments, and we need to be able to do repeated experiments. Today we don’t really have a way to actually monitor it, to really do experiments, because what happens with this input to the brain is a very important thing, and in the next couple of years more people should devote themselves to brain science.” Shen xiangyang explained.

Therefore, Li thinks that machine tools in the future should learn human intentions, hoping to solve natural language with artificial intelligence. Represents the biggest direction of artificial intelligence development.

In fact, any problem that can be solved with technology is not a problem. What worries most about AI is what will happen to humans if the singularity of AI arrives.

Giiso information, founded in 2013, is the first domestic high-tech enterprise focusing on the research and development of intelligent information processing technology and the development and operation of core software for writing robots. At the beginning of its establishment, the company received angel round investment, and in August 2015, GSR Venture Capital received $5 million pre-A round of investment.

Zhu min proposed ten ethical challenges that AI may encounter, such as career problems caused by ai replacing human beings; How to distribute the wealth produced by machines; How machines change human behavior and interaction; How to eliminate machine bias; How to ensure that the machine does not produce destructive effect; How to prevent the machine from making mistakes; How to prevent machines from becoming weapons of attack; The invisible machine itself is an ethical issue; What about machines that enhance human functions? How do humans control complex artificial intelligence systems? Robot rights, should intelligent robots have human rights?