AI four major schools Huashan sword, who can dominate the world?

In 2016, Google’s victory with AlphaGo revealed its ambition to occupy the AI ​​market. Facebook plans its own development route for artificial intelligence and drones in the next 10 years. Domestic BAT is also gearing up in the AI ​​field. Internet giants are rushing to the field of artificial intelligence, and some niche groups or computer professionals are not far behind.

Previously, the data 猿 reporter was invited to participate in the “First Shanghai BOT Big Data Application Contest”. In the first line, I deeply felt that the ideas of talents in various fields for AI innovation applications are endless. Many university institutions and entrepreneurs are developing AI industry. The enthusiasm continues to climb.

Obviously, artificial intelligence and machine learning are important new technologies, and everyone is trying to board this train to the future AI. However, if the industry is only a single company, the development of AI technology is bound to become very slow and inefficient. Therefore, the greater the competition and the greater the competition, the better the development of AI.

AI four major schools Huashan sword, who can dominate the world?

From the perspective of the global market, journalists believe that AI companies can be divided into four major factions.

First, the technical school "big cow"

AI four major schools Huashan sword, who can dominate the world?

I have to say that at the beginning of this year, the wonderful Go game between AlphaGo and Li Shishi was the "originator" of this artificial intelligence boom. In various media reports and information bombings, international giants such as Google, Microsoft and Facebook have been at the forefront of the AI ​​corporate rankings.

On the whole, the AI ​​technology companies represented by Google have three characteristics:

First, it has powerful AI technology;

Second, have a large number of top AI talents;

Third, we have constantly updated high-tech products.

Of course, with these strong advantages, the most crucial thing is to have strong financial support.

From a technical perspective, in November 2015, Google opened up the second generation deep learning system Tensorflow. The system can be applied to many fields such as speech recognition and photo recognition, and it is faster and more flexible. In May of this year, at the annual Google I/O conference, Google Virtual Assistant, Google Assistant, made its debut. In terms of cloud services, Google unifies the "cloud vision" application interface. The same system recognizes logos, landmarks, faces, and optical character recognition of texts. These systems use their own developed Tensor processor (TPU). ).

From the perspective of talents, as more and more companies integrate AI technology into products and services, the scarce AI talents have become more and more darlings. On November 22, it was reported that Google will invest more than $3.4 million to support seven researchers from the Montreal Algorithm Research Laboratory in Canada to conduct AI product development, and will also set up a new deep learning and AI research team. And this is not Google's most sensational AI talent reserve action.

As early as January 2014, Google spent $660 million to acquire the UK's DeepMind artificial intelligence research team, which is the most important behind-the-scenes of AlphaGo. DeepMind focuses on deep learning and intensive learning research, allowing the machine to do a lot of learning on its own, and DeepMind and Google Brain (Google's dedicated AI department in 2011) often interact, which is called the right arm of Google in the development of artificial intelligence.

In fact, in addition to Google, large companies such as Microsoft, Facebook and Apple have opened AI research centers around the world, and they have top talents. For example, Facebook's application machine learning team, its primary goal is to apply AI technology to Facebook products; and one-fifth of the entire Facebook team belongs to the team.

In short, these technology giants have an unassailable leading position in terms of both technical aspects and talent reserves.

Second, implicit pragmatism

Compared with Google's high-profile and stride forward artificial intelligence development strategy, there are still some companies that are steadily moving in a low-key manner. Such as Microsoft and Baidu.

Twenty-five years ago, Bill Gates founded Microsoft Research, whose main focus was on technical research in speech recognition, natural language, and computer vision. Through years of investment and in-depth understanding of these areas, Microsoft Skype instant translation, Xiaobing chat robot and Xiaona virtual assistant appeared one after another. Today, Xiaona is serving 113 million users every day, and has answered more than 12 billion questions. She has a very large corpus database; Xiao Bing is no longer a purely “entertainment product”, but is relying on technological evolution to approach nature. Interact to achieve true artificial intelligence.

In September of this year, Microsoft has integrated more than 5,000 computer scientists and engineers including Microsoft Research, Microsoft Information Platform, Bing and Xiaona Products, and established a new artificial intelligence and R&D department. In the future, we will focus on injecting AI technology into Microsoft products and services.

In China, Baidu's investment and research and development in the field of artificial intelligence is particularly eye-catching. In April this year, innovative businesses such as unmanned vehicles and artificial intelligence were independently isolated from Baidu and directly managed by Li Yanhong. The high-level representatives represented by Li Yanhong have repeatedly said in public that artificial intelligence will be the top priority of Baidu's next development. Of course, Baidu's artificial intelligence is not just in the laboratory stage. The layout in O2O, unmanned vehicles, smart cities and other fields can be described as full bloom.

At present, Baidu has embedded voice and image technology into mobile phone Baidu, Baidu map, Baidu input method, Baidu picture, degree secret and other star products. In December last year, Baidu released an unmanned vehicle. In less than a year, it established a number of unmanned vehicle demonstration zones, and simultaneously tested roads in both China and the United States, and collected massive road test data to train Baidu's brain. Perhaps in this round of artificial intelligence boom, Baidu will dominate domestic AI development and even rewrite the domestic Internet BAT pattern, because in comparison, Ali is still mainly cooperating with its own cloud computing, big data technology for e-commerce platform and Internet of Things. Develop artificial intelligence.

Third, the hardware sent "new people"

With the explosion of massive data, the application of artificial intelligence relies on two major factors: algorithm and data. It is not difficult to imagine that when the international giants such as Google, Facebook, and Microsoft are involved in the field of artificial intelligence, the hardware requirements and demands of various enterprises for AI development must be very urgent. The next one to stand in the limelight is the hardware vendor.

In the current market, many technology giants starting from hardware are quietly turning to artificial intelligence with their excellent hardware systems. Taking Intel as an example, its entry into the field of artificial intelligence is a natural evolution of technology and a natural choice to adapt to market changes.

As we all know, the key technology of artificial intelligence - neural network involves very complex algorithms. Currently, almost all companies developing AI are using GPUs for deep neural network training. GPUs are targeted at dense, high-parallel computing, with more processing units and more transistors for data processing. With the development of image speech recognition, driverless and other application layers, GPU is rapidly expanding its market share.

At present, there are many manufacturers of GPUs, but the only ones that are familiar with them are "INA", namely Intel, NVIDIA and AMD. If calculated according to market share, Intel's integrated GPUs sold with motherboards and CPUs already account for more than 60% of the entire market. According to some data, nearly 7% of all servers in the world are running load related to machine learning, and most of them use Intel processors.

However, the high demand for computational resources in deep learning is like a technical black hole, and the giants are more inclined to use more specialized chips.

In December 2015, Intel acquired Altera, the world's second largest FPGA vendor, for the largest $16.7 billion in history. FPGA is a technology with certain programmability between dedicated chip and general-purpose chip, which can perform data parallel and task parallel computing at the same time; when processing a specific application, its energy consumption ratio is even 10 times that of CPU, GPU 3 times. At present, FPGA has been applied in many fields such as logic control, signal processing, and image processing. Intel's move is also considered to be another important strategy for deploying artificial intelligence.

Not long ago, Intel also showed the public a new artificial intelligence processor code-named "Nervana", and announced that the relevant model will be tested in the middle of next year. If it goes smoothly, the final chip will be launched by the end of 2017.

In contrast, the technology-based Google is also developing TPU chips for artificial intelligence products, and is regarded by the industry as a GPU competitor, but it is only used in the internal data center of the enterprise, and cannot be used as a commercial product. IBM's True North and other vendors' similar chips are also emerging, but have not yet been introduced to the market.

In general, Internet companies with hardware advantages are bound to gain a share in the future AI market.

Fourth, the small start-up

The concept of artificial intelligence has been well known to the public, and various industries have also promoted it as the ultimate application goal. However, as far as the current level of application is concerned, the most mature and popular ones should be personal assistants, such as Apple Siri, Microsoft Xiaona, and Google's Google Now. There are the Confucius of Science and Technology, Turing's wormhole voice assistant, etc. Assistants generally use a PC or mobile phone.

With the development of technology, artificial intelligence robots have begun to gradually integrate realistic scenes in addition to voice functions. In April 2016, “Xian Er Machine” was unexpectedly popular with funny interaction and ingenious language. This robot is jointly launched by Longquan Temple and Haizhi Intelligent Company to promote Dharma. Among them, Longquan Temple provides Dharma knowledge content, and Haizhi Intelligent provides algorithm technology. Artificial intelligence robots such as these are coming into the market and showing their edge.

Previously, in the "First Shanghai BOT Contest", the popularity of chat robots far exceeded the reporter's expectations. In addition to the familiar financial fields, medical, education, law, automotive and other fields are rushing to get involved in artificial intelligence technology. As if overnight, artificial intelligence has become the ultimate standard in all industries.

However, the threshold of AI technology is well known. It is a comprehensive skill that combines big data, professional knowledge and technical algorithms in various industries. In the current market, most of the AI-type startups are only deeply ploughed in a certain field, and the product homogenization phenomenon is serious. Basically, only one of them is acquired, and the so-called intelligent robot made is not perfect artificial intelligence. Artificial intelligence should be more aesthetic and more spiritual than the robots you and I envision.

2016 is the first year of the explosion of artificial intelligence. All parties are working hard to compete for the future market of AI. The competition in the industry has been like an arrow. Who can become an industry giant, there is no trace. I hope that we will sink into the current AI boom, but also calmly think about how to realize the commercialization of AI in the future, how to establish the technical barriers of the enterprise itself, rather than self-righteous biased artificial intelligence value!

Ground Screw For Other Buildings

High Dip Galvanized Ground Screw Pile made of carbon steel :Q235 ,according to standard ;DIN EN ISO1461-1999 ,not only used on the Solar Mounting projects,but also can be use for many other fields ,such as some temporary building,for timber,flage and banner,traffice sign ,for fencing construction ,for container house,for vineyard fencing ,for bridge & . it is easy to install ,no need digging ,no-concrete ,save time and cost .no harm to environment .

Due to their design and ease to installation,they are most commonly used wheneversoil conditons preventstandard foundation solutions,instead of requiring large excavationwork,they thread into the ground.

Construction Ground Screw,Hdg Ground Screw,Screw Anchor,Helical Pier,Ground Screw Anchor


This entry was posted in on