Foreign media commentary on Li Kaifu’s “AI future”: the four wave is sweeping the globe.
【Li Kaifu is one of the world’s most watched AI investors, managing more than $2 billion between six funds in the United States and more than 300 portfolio companies in China.
With his pioneering work in artificial intelligence, executive leadership of Microsoft, Apple and Google (he was the founding president of Google China), and his venture capital fund, Sinovation Ventures, Lee shares the following valuable insights:
1、Four factors that drive today’s AI ecosystem.
2、China’s extraordinary progress in AI applications;
3、The trend of autonomous system;
4、How are we going to adapt?
Based on Beijing and Silicon Valley, Mr. Kaifu Lee examines the Chinese and American technology giants. In this article, I will discuss the “four waves of the development of artificial intelligence” mentioned by Kaifu Lee, which is a good framework for discussing the current situation and development direction of artificial intelligence. I will also introduce some of the hottest Chinese branches.Technology companies, these companies deserve attention.
The first wave: Internet AI
In the first phase of AI deployment, we mainly deal with recommendation engine algorithm systems, which learn from a large number of user data to plan our online content for each of us.
Think about Amazon’s spot recommendation, or the “Up Next” YouTube video you have to watch before you get back to work, or the Facebook ad that seems to know what you’re going to buy before you buy it.
Internet AI takes advantage of the fact that users automatically tag data as they browse through our network. Click on it or stay on one page longer than we do on another, or stay on a video on FacebookLet’s see what will happen in the end.
These tagged data cascades create a detailed portrait of our personalities, habits, needs, and desires: a “perfect recipe” for getting customized content on a specific platform.
Currently, Lee estimates that Chinese and American companies face direct competition in deploying Internet AI. But given China’s data advantages, he expects China’s technology giants to be slightly ahead of their US counterparts in the next five years.
Although you have heard of Alibaba and Baidu, you may never meet headlines. Starting with imitating the popular Buzz feed in the U.S., the headline was valued at $20 billion in 2017, making Buzzfeed more than 10 times its value. However, with nearly 1 per day.2 billion active users, headlines do not stop creating viral content.
The headline AI engine, equipped with natural language processing and computer vision, surveys the vast network of different websites and contributors, rewrites headlines to optimize user participation, and handles each user’s online behavior — click, comment, participation time — to provide personalized news to millions of consumersSubscribe. As users increasingly use headline content, the company’s algorithms are getting better and better at recommending content, optimizing headlines, and providing truly personalized information flows.
It is this positive feedback loop that drives the impact of today’s AI giants on the Internet AI wave.
The second wave: commercial AI
Although Internet AI uses users to constantly mark data by clicking and other participating indicators, commercial AI skips data that traditional companies have labeled in the past.
Think about the ratio of loans made and repayments recorded by banks; the hospital archives diagnostic, imaging data and subsequent health outcomes; or the court records conviction history, recidivism, and escape. Although we humans make predictions based on obvious root causes (strong characteristics), artificial intelligence algorithms can handle thousands of weakly correlated variables (weak characteristics).These variables may have more relationships with the given results, rather than the usual guesses. By finding hidden correlations and avoiding our linear causal logic, commercial AI uses tag data to train algorithms that can even outperform the most experienced experts. These data trained AI engines are applied toIn banking, insurance and legal decisions, you can minimize default rates, optimize premiums and recidivism rates.
Although Lee confidently puts the U.S. ahead of commercial AI, China’s huge lag in structured industry data could actually benefit the future. In China, start-ups can skip the traditional system of industries, and China has a great advantage. Take China’s intelligent financial application as an example. 70s of last centuryWhen the Americans embraced credit cards and debit cards, China missed this technology. Fast forward to 2017, China’s mobile payments outnumber Americans, a ratio of 50 to 1. In the absence of entrenched credit card competition, mobile payment is clearly a rise in China’s cash-intensive economy.Grade. By the end of 2017, 70% of China’s 753 million smart phone users had accepted mobile payment. However, across credit cards and mobile payments, China has largely abandoned the concept of credit. This is the role of intelligent finance.
As an AI application for microfinance, smart finance relies almost entirely on its algorithms to make millions of microloans. For each potential borrower, the app simply requires access to a part of the user’s mobile data. According to your typing speed and the percentage change of battery,Smart financial applications can predict your repayment probability of $300 with astonishing accuracy.
The deployment of commercial AI and Internet AI has completely changed our industry and personal lifestyle. However, there are two more waves on the horizon — the perception of AI and the autonomous AI.
The third wave: perception AI
In this wave, artificial intelligence will be upgraded through the eyes, ears and countless other senses, combining the digital world with our physical environment.
As sensors and smart devices proliferate in our homes and cities, we are entering a trillion sensor economy.
Companies like China Millet are rolling out millions of Internet of Things devices, and research teams have begun developing smart dust prototypes, solar cells and sensor particles, that can store and exchange data anytime, anywhere.
As Lee explains, perceiving AI “will bring convenience and richness into our offline reality.” Sensor-supported hardware will turn everything from hospitals to cars to schools into online-merge-offliesNE) environment.
Imagine walking into a department store, recognizing faces, picking up the items you buy most often, and pulling up a virtual assistant (VA) cart. After you have pre-loaded your data, the cart adjusts your shopping list and provides voice input to remind you of the upcoming anniversary.Buy your spouse’s favorite wine and guide you through personalized shop routes.
Although we haven’t taken full advantage of the full potential of perceiving AI, China and the United States have made incredible progress.
As I discussed last week, Shenzhen has become a powerful push for smart hardware as a base for robots, unmanned aerial vehicles, wearable technology, and Internet of Things infrastructure start-ups. With thousands of factories exporting sensors and electronic parts, Shenzhen’s skilled engineers have unprecedented disciplineThe prototype and speed are prototyping and iterating over the original product.
In this wave, companies like millet aim to turn bathrooms, kitchens and living rooms into smart OMO environments. After investing in 220 companies and incubating 29 start-ups that produce their products, millet had more than 85 million networking devices by the end of 2017, becomingThe largest product connection network in the world.
In China, a KFC restaurant has even launched the “pay-with-your-face” function with Alipay (Alibaba’s mobile payment platform). Forget about cash, credit cards and mobile phones. Let OMO do the job.
The fourth wave: automated AI (Autonomous AI)
The most memorable — and unpredictable — was the fourth and last wave: automated AI. Integrating all previous developments, automated AI allows machines to perceive and respond to the world around them, moving and acting effectively.
Although today’s machines can do better in structured or even unstructured environments, such as the Atlas robot of Boston power company or the upcoming autopilot, the machine with vision, hearing, touch and data optimization will be a brand new picture.
Imagine: a swarm of UAVs can selectively spray and harvest the entire farm with computer vision and extraordinary dexterity. Heat-resistant UAVs can extinguish forest fires more effectively, and L5 level self driving cars can independently and intelligently drive on roads and transportation systems.
Although automated AI involves, first and foremost, robots that create immediate economic value — automating tasks on a one-to-one basis — these smart machines will eventually revolutionize the industry.
Li Kaifu believes that the United States has the leading edge in the field of automatic AI, especially in the aspect of automatic driving vehicles. But the Chinese government’s efforts are rapidly increasing competition.
In Zhejiang Province, China, highway regulators and government officials have planned to build China’s first smart highway, equipped with sensors, embedded solar panels on roads, and wireless communications between cars, roads and drivers. The project aims to increase the transport efficiency by 30% while minimizing the number of deaths.One day, it may be possible to support the continuous charging of automatic electric vehicles during driving.
A similar government promotion project involves Beijing’s new neighbour, Xiong an. In the next 20 years, Xiong an new district is expected to invest more than $580 billion in infrastructure spending. It may one day become the first city in the world to build a self driving car. Baidu is already working with the local government of Xiong an.To build the AI city in an environmentally friendly way.
Finally, Li Kaifu predicts that China will almost certainly take the lead in the field of UAVs. Shenzhen has become the leading position of the UAV manufacturer Dajiang. Called “the best company I’ve ever met” by Chris Anderson, Ozawa owns 50% of North America.The UAV market has been driven by excellent manufacturers in Shenzhen.
One thing is certain, though, in the fourth wave: in decades, we will witness the emergence of artificial intelligence embedded in urban landscapes and autonomous robots that can interact with the real world to help address today’s most pressing challenges.
Selected from: Singularityhub
Authors: Peter H. Diamandis
Compile: NetEase intelligence