At present, artificial intelligence knowledge has been spread among Chinese enterprises, helping Chinese enterprises to move towards the era of artificial intelligence. In addition, many young people are actively involved in artificial intelligence, and the number of growth rate beyond imagination.
In his speech, Kai-fu Lee also mentioned China's mobile payments. At present, Chinese people use mobile phones to pay 50 times more frequently than Americans when shopping. China has more than 10 times the take-out delivery volume of the United States, and mobile services have infiltrated people's online and offline lives. They are growing faster in China than the United States.
In terms of innovation, Kai-fu Lee said Copy to China has become a Copy from China. Many foreign products began to learn from China's characteristics and inspiration.
Kai-fu Lee said that the overall thinking of China is a bold attempt and a rapid iteration. If problems do arise, they will not be totally denied and a solution to the problem will be found. We will get more and more data, AI performance will be getting better and better, will eventually promote China's technological progress and become an artificial intelligence powerhouse. (Zhao Wentao)
Attached speech full text:
On the way to the meeting this morning, the driver of Uber asked me what I would like to attend and I said I would like to go to MIT to discuss the future AI to replace human work.
He asked me, "How long will the driverless car go to market? By then I will be out of work."
I glanced at him and said, "This is the United States. It will be almost 15 to 20 years later."
He relieved, "Well, I'll retire then."
This dialogue did not happen in China. If you ask me in China, I may have to say about 10 years because the development of AI in China will be faster and more efficient.
This also shows that I am optimistic about the future development of artificial intelligence in China.
Next, I will talk about the current situation and future trends in the development of artificial intelligence in China. I will try my best to cover the aspects of artificial intelligence technology, market, products, funds and policies.
First, technology: AI talent in China is growing at an incredible rate
This is not a technical talk, but we usually have to talk about technology first.
Obviously, AI Turing Award winners, deep learning explorers, not many Chinese people, the elite of the AI team, senior team and super team leaders are Americans.
However, I want to tell you that although the top teams are led by the Americans, when we look at China, we can see that the Chinese AI team is growing at an incredible rate.
In 1998, I createdMicrosoftChina Academy (later upgraded to "Microsoft Asia Research Institute"), so far, Microsoft Asia Research Institute has trained 5,000 people, the main job is to do artificial intelligence research and development.
The nine people listed on the left side of the PPT, all from Microsoft Research Asia, are now Baidu, Alibaba,Tencent, Chief technology officer of today's headlines and Haier, or head of artificial intelligence.
In general, knowledge of artificial intelligence has been disseminated among Chinese enterprises, helping Chinese enterprises to move toward artificial intelligence. In addition, many young people are actively involved in artificial intelligence, and the number of growth rate beyond imagination.
The technology of Chinese enterprises is really backward, but these enterprises are trying hard to catch up and actively join the army of technological development.
Recently, Face ++, a company invested by Innovation Works, defeated the top enterprise teams from Microsoft, Facebook and Google in the MS COCO competition, the most watched and authoritative competition in the field of computer vision. It won three first-first The second result, becoming the first Chinese company to win the competition.
Now, the company is working to identify three million faces at any given time, a technology widely used in security and business activities around the world, and Apple's new machineiPhoneX designed this feature.
At present, many major handset manufacturers in China are struggling to catch up with the iPhone. I can not pinpoint who they all have, but most do that and mostly choose to use Face ++ technology.
As we all know, technology does not have to be published in the paper, but it requires a lot of data and tedious work, the contents of the paper applied to the actual situation.
In the past, the number of AI thesis authors in China was very small. However, from 2006 to 2015, the top AI papers that Chinese authors participated in, accounting for the proportion of all top AI papers, increased from 23.2% to 42.8% year by year. The number of citations of top AI papers participated by Chinese authors has increased from 25.5% to 55.8% of the quotations of all the top-level AI papers.
Second, the market: China Mobile pay 50 times higher than the Americans
The size of the data is very important for artificial intelligence algorithms, the more data, the better the things.
If you have more data, products upgraded through artificial intelligence will help you get more users and help you make more money.
After that, you can hire more scientists, buy more machines and get more data. This cycle is a key success factor for Google, Facebook, Microsoft, and China's Baidu, Tencent and Alibaba and many others.
So, before that the top scientists can make a miracle, this is not the case. No matter at any moment, a brilliant scientist with a lot of data will eventually beat top scientists who have a small amount of data.
So how much data does China have? We all know that China is the largest user base in the global handset and Internet markets. You may think of drops, Taobao, Meituan these applications, they have a huge amount of users.
To give you an example, China's mobile payment.
The United States has the world's best payment system - credit cards, and China is still lagging behind so far, with credit-card penetration of only about 20%. But now, China Mobile Payments have basically been implemented everywhere.
What is mobile payment? First, unlike credit cards, unlike banks, handling fees are basically free. Second, it uses end-to-end technology where anyone can transfer money to anyone else, not necessarily a merchant or consumer. Third, it achieves real-time arrival without the need for approval of credit card transfers.
Chinese use mobile phones to pay 50 times more frequently when shopping than Americans. China has more than 10 times the take-out delivery volume of the United States, and mobile services have infiltrated people's online and offline lives. They are growing faster in China than the United States.
If you really go to China and spend a day with a Chinese, what they are doing will make you very surprised.
They are out of cash now, and they pay for everything by their mobile phone. When eating, do not have to go to restaurants, takeout can be set, takeaway brother will be on time for your order. China has also seen the sharing of bicycles, both for users to create value, but also reduce vehicle emissions, make the environment a better place. These are the changes that mobile applications bring to Chinese life.
What will happen next is - online and offline integration, OMO (Online-Merge-Offline) era.
Currently, many companies are able to track offline users and transmit their behavior online, fully integrated with online activities.
In more detail, when you are thereAmazonWhen browsing the information, Amazon knows which page you're surfing, what you clicked on, what you bought, what you did not buy, who you were, and what you bought before.
In the future, all this will happen in real life, with cameras and sensors used in stores and shopping malls, knowing in real time what you are doing and transferring information to the dataserver, Make more accurate predictions about what you want to do.
Stores can also use this information to source the right inventory, create the right supply chain, and design the product to meet the needs of the user, thus linking all aspects.
This situation can occur not only in the consumer industry but also in education and many other areas.
These changes are quite impressive, and while AmazonGo is technically the most advanced, self-service stores across China are also evolving. If you go to China in another year, you will find that half of the stores have been semi-automatically or fully automated. This is the efficiency of Chinese enterprises.
Third, the product: the market value of Internet companies in China and the United States close to 1: 1
In terms of product? China is able to develop high-quality artificial intelligence products?
In the majority of people's minds, China has been stuffing others and has not had its own innovation, but it was already 15 years ago.
Imitation is not a misappropriation of other people's intellectual property, but to imitate the appearance of other products, features. It is through imitation that entrepreneurs have learned how to develop products that meet the needs of their users and then become good product stewards and quickly move on to the next stage of development.
At this stage, based on the innovative skills abroad, Chinese entrepreneurs have developed products with unique characteristics. For example, microblogging, Taobao, although they are not original, but based on a large number of market feedback developed faster, and has gone beyond the respective fields of benchmarking products: Twitter and eBay.
Later, Chinese companies started to research and launch new and innovative products that the United States did not have. When I was working at Google, we spent many years researching customized news, but today the headlines got it done earlier than Google did. Sharing bicycles now is a world-leading business model.
After that, Copy to China became a Copy from China. Many foreign products began to learn from China's characteristics and inspiration. So it is not surprising to say that Chinese companies have now caught up with the American business.
China's artificial intelligence products will achieve leaps and bounds, beyond the United States products? Here I will make some predictions. This prediction combines the four waves of artificial intelligence.
Innovation Workshop believes that the arrival of artificial intelligence has four waves.
The first wave is the Internet intelligent, by calling the Internet a lot of data, so that Internet sites, APP become more intelligent;
The second wave of business intelligence is the activation of many years of data retained by commercial companies and the resulting commercial value;
The third wave is the intellectualization of the physical world, digitizing many of the past non-digital behaviors in the real world;
The fourth wave is fully automated intelligent, driverless, robotic, etc. into business applications.
In the first wave, through the use of Internet data, China's Internet companies will achieve leaps and bounds, surpassing the U.S. Internet companies. The reason is that everybody has a cell phone and they pay for it, and new and innovative products will emerge, but not every single company will surpass.
In the second wave, as Chinese companies are not good at creating databases, Chinese companies lag far behind US businesses in terms of business data.
For these enterprises, the use of data mining is a very difficult task, because to achieve data structure takes some time. But now the government has begun to support them and has set preferential policies to promote enterprises to accelerate data structure. Chinese companies may catch up with US companies in this area in the future, but it is not yet possible.
In the third wave, China will achieve leapfrog development and overtake the United States. To share a bicycle, for example, riding a bike, sharing a lot of GPS data transmission. The transmission of these data depends on a large number of sensors. China's hardware prices are much lower than the United States, for example, similar to the Amazon Smart Speaker Lynx Elf, in China, the promotional price of only 15 US dollars.
In the fourth wave, the chances of China and the United States are each 50%. The United States will take the lead in driverless cars and robotics and will continue to lead in the next two years. However, with the support of the government, Chinese enterprises will also achieve faster development.
Fourth, funding: a large number of financial inflows artificial intelligence
What kind of money is it? In short, a lot of money flows into the AI business. Last year, there was also a clear and significant increase in the Chinese government's Matching Funds.
Government-led funds refer to the government as a limited partner in venture capital funds and willing to earn less than other limited partners to help top venture capital funds spend less time to raise funds and have more time to invest . Inspired by Israel and Singapore, this approach is a very wise move.
In the secondary market, artificial intelligence businessstockPrices, ushered in a crazy soaring trend. From China, IETF is a speech recognition company that in many ways matches the Silicon Valley company Nuance. At present, the gap between the two companies in the capital market has reached 6 billion U.S. dollars, up from a year earlier. The large inflow of funds, although it also caused some bubbles, stimulated more young talents to enter the field of artificial intelligence.
V. Policy: The spirit of opening up will create an AI powerhouse
The last point, policy. The most important policy in the field of AI in China is the "New Generation of Artificial Intelligence Development Plan" promulgated in July this year. This report is very straightforward. By 2020, artificial intelligence technology will be synchronized with the world advanced level. By 2030, it will become the world's major Artificial Intelligence Innovation Center.
The implementation of China's policy is very high. For example, in 2010, China proposed to be a leader in the high-speed rail industry. Many people are skeptical of this. But six years later, China owned 60% of the world's high-speed rail. Therefore, we can expect that the plan of artificial intelligence in China will soon become a reality.
The Chinese government supports technological development. When new technologies emerge, the government will certainly support it, unlike the policy that the United States is good at forbidding.
China's overall thinking is a bold attempt and a rapid iteration. If problems do arise, they will not be totally discarded and solutions will be found to be a positive solution. We will get more and more data, AI performance will be getting better and better, will eventually promote China's technological progress and become an artificial intelligence powerhouse.