For a long time, Google has been using AI technology in its core products and constantly developing new AI products. So far, Google’s achievements in AI have given a strong impression – Google has had a good time. In addition to DeepMind's strides, the creative Google headquarters for the development of AI also presents a scene of blooming everywhere.
This year, Google CEO SundarPichai announced that Google will move from MobileFirst (Mobile First) to AIFirst (Artificial Intelligence First). In fact, Google has long used AI (artificial intelligence) and machine learning technology (one of the core technologies of artificial intelligence) in its core products. Not only that, but it also acquired other AI companies and continuously developed new AI. product. So far, Google’s achievements in AI have given a strong impression – Google has had a great time!
Acquisition of DeepMind: A great value business Why does Google make people feel that it is having a great time? This may be due to the acquisition of the artificial intelligence company DeepMind. DeepMind, co-founded in 2010 by artificial intelligence programmer and neuroscientist Demis Hassabis, and in 2014, Google spent £400 million (converted at $600 million at the exchange rate) to acquire it. In the past two years, the international artificial intelligence boom has been ignited. Thanks to DeepMind for its outstanding performance:
In 2014, DeepMind developed AlphaGo5:0 to defeat the European Go champion Fan Wei;
In March of this year, AlphaGo won the former world number one Li Shishi with a score of 4:1;
On November 7th, Fan Wei, who has joined the AlphaGo team, announced that AlphaGo’s chess power has soared and that he will return to play next year. Whether we will fight China’s Ke Jie, we will wait and see.
Earlier this month, Blizzard Carnival reported that DeepMind will work with Blizzard (American game developer) to develop tools on DeepMind's platform to let third parties teach artificial intelligence to play real-time game StarCraft II.
DeepMind seems to have been teaching AI to play games and is playing very well. Of course, DeepMind is a company acquired by Google. It is not Google's headquarters. But Google spent 400 million pounds to buy it. Its research trend is unlikely to be unaffected by the headquarters. If the headquarters does not agree, how can DeepMind play so happy? In fact, DeepMind appears to be playing games on the surface, but it is essentially a unique way of technology development.
At present, AI is still in the stage of weak artificial intelligence. There are still many problems to be solved in order to invest in deeper and wider practical applications. It can be said that this is not only an area with unlimited potential, but also a field full of challenges and a large number of super brains to overcome. This year, DemisHassabis said in an interview with TheVerge, a US technology media website, that DeepMind's goal is not only to win in the game, but also to have fun and inspiration. Playing games is actually another test bed - a platform for trying to write algorithms and test algorithms. In the end, I hope to apply technology to solve real-world problems.
Hassabis hopes to create a "comprehensive intelligence" that can do any task through self-learning like humans. Previously playing the first-class AlphaGo in the world of Go, there is still a lot of obstacles to cross the "integrated intelligence" - it must first master the messy real world. So the DeepMind team thought of StarCraft. This game is an interesting test environment for current AI research, and it provides a useful bridge to the real world. In other words, DeepMind takes this game as an important transition for the AI ​​to adapt to the real world, because the complexity of StarCraft is closer to the real world than to chess, even for the most powerful computers. The system is still a serious challenge. The insights gained from StarCraft can make artificial intelligence process problems faster and more efficiently than before. On the other hand, chess is played in turn, and StarCraft is launched in real time. That's why we saw Google's artificial intelligence going to play StarCraft II.
For DeepMind, Google's support is very important. Without Google, there is no today's DeepMind, they are a win-win situation. -- DeepMind, which was acquired by Google, has grown rapidly, adding dozens of researchers. In October 2015, it acquired two companies created by Oxford professors, DarkBlueLabs (Deep Blue Labs) and VisionFactory (Vision Factory). . In terms of optimizing research progress, Google gave DeepMind considerable autonomy; on the other hand, DeepMind needed to train AlphaGo with a lot of hardware, and also developed different versions to play in the cloud. Without the resources provided by Google, DeepMind could not Progress is made in the short term.
Google's support for DeepMind is because Google wants to solve all other problems by solving the problem of artificial intelligence. DeepMind is one of the most powerful right-handers of Google's development of artificial intelligence. In fact, in addition to its own independent research, DeepMind also participates in many product developments within Google. Hassabis said that DeepMind and Google Brain (the AI ​​project born in GoogleX Lab 2011) interact weekly. DeepMind focuses on enhancing learning, while Google Brain is more focused on deep learning.
Said that the acquisition of DeepMind is a good value business, you have to mention the other outstanding performance of DeepMind, this is indeed a technically leading artificial intelligence company:
In July of this year, Hassabis said that DeepMind's AI system is helping Google save huge electricity bills. The system reduces power consumption by controlling some of the data center's equipment—manipulating servers and cooling systems. Google also said that the technology does increase power usage efficiency (PUE) by 15%. Google’s electricity consumption in 2014 was 4,402,836 MWh, and a large portion of the electricity was consumed by the data center. According to the US Energy Information Administration, the price of electricity paid by enterprises is about 25~40 US dollars per MWh. No matter which standard is calculated, if the power consumption of the data center is reduced by 10%, it may be Google for several years. Save hundreds of millions of dollars in electricity bills (so that Google’s $600 million in the purchase of DeepMind will soon be “saving†back). Hassabis said: "The AI ​​system controls 120 variables in the data center, including fans, cooling systems and windows." DeepMind may ask Google to add sensors to the data center in the future to further save power through richer information.
Google's artificial intelligence road: injecting vitality into AI with strong creativity
Founded in 1998, Google is recognized as the world's largest search engine company. The multinational technology company announced in August 2015 that it has adjusted its corporate structure and created an umbrella company called Alphabet. The subsidiary of Alphabet has been transformed into a high-tech enterprise covering a wide range of fields.This year, Google CEO SundarPichai announced that Google will move from MobileFirst (Mobile First) to AIFirst (Artificial Intelligence First). In fact, Google has long used AI (artificial intelligence) and machine learning technology (one of the core technologies of artificial intelligence) in its core products. Not only that, but it also acquired other AI companies and continuously developed new AI. product. So far, Google’s achievements in AI have given a strong impression – Google has had a great time!
In 2014, DeepMind developed AlphaGo5:0 to defeat the European Go champion Fan Wei;
In March of this year, AlphaGo won the former world number one Li Shishi with a score of 4:1;
On November 7th, Fan Wei, who has joined the AlphaGo team, announced that AlphaGo’s chess power has soared and that he will return to play next year. Whether we will fight China’s Ke Jie, we will wait and see.
Earlier this month, Blizzard Carnival reported that DeepMind will work with Blizzard (American game developer) to develop tools on DeepMind's platform to let third parties teach artificial intelligence to play real-time game StarCraft II.
DeepMind seems to have been teaching AI to play games and is playing very well. Of course, DeepMind is a company acquired by Google. It is not Google's headquarters. But Google spent 400 million pounds to buy it. Its research trend is unlikely to be unaffected by the headquarters. If the headquarters does not agree, how can DeepMind play so happy? In fact, DeepMind appears to be playing games on the surface, but it is essentially a unique way of technology development.
At present, AI is still in the stage of weak artificial intelligence. There are still many problems to be solved in order to invest in deeper and wider practical applications. It can be said that this is not only an area with unlimited potential, but also a field full of challenges and a large number of super brains to overcome. This year, DemisHassabis said in an interview with TheVerge, a US technology media website, that DeepMind's goal is not only to win in the game, but also to have fun and inspiration. Playing games is actually another test bed - a platform for trying to write algorithms and test algorithms. In the end, I hope to apply technology to solve real-world problems.
Hassabis hopes to create a "comprehensive intelligence" that can do any task through self-learning like humans. Previously playing the first-class AlphaGo in the world of Go, there is still a lot of obstacles to cross the "integrated intelligence" - it must first master the messy real world. So the DeepMind team thought of StarCraft. This game is an interesting test environment for current AI research, and it provides a useful bridge to the real world. In other words, DeepMind takes this game as an important transition for the AI ​​to adapt to the real world, because the complexity of StarCraft is closer to the real world than to chess, even for the most powerful computers. The system is still a serious challenge. The insights gained from StarCraft can make artificial intelligence process problems faster and more efficiently than before. On the other hand, chess is played in turn, and StarCraft is launched in real time. That's why we saw Google's artificial intelligence going to play StarCraft II.
For DeepMind, Google's support is very important. Without Google, there is no today's DeepMind, they are a win-win situation. -- DeepMind, which was acquired by Google, has grown rapidly, adding dozens of researchers. In October 2015, it acquired two companies created by Oxford professors, DarkBlueLabs (Deep Blue Labs) and VisionFactory (Vision Factory). . In terms of optimizing research progress, Google gave DeepMind considerable autonomy; on the other hand, DeepMind needed to train AlphaGo with a lot of hardware, and also developed different versions to play in the cloud. Without the resources provided by Google, DeepMind could not Progress is made in the short term.
Google's support for DeepMind is because Google wants to solve all other problems by solving the problem of artificial intelligence. DeepMind is one of the most powerful right-handers of Google's development of artificial intelligence. In fact, in addition to its own independent research, DeepMind also participates in many product developments within Google. Hassabis said that DeepMind and Google Brain (the AI ​​project born in GoogleX Lab 2011) interact weekly. DeepMind focuses on enhancing learning, while Google Brain is more focused on deep learning.
Said that the acquisition of DeepMind is a good value business, you have to mention the other outstanding performance of DeepMind, this is indeed a technically leading artificial intelligence company:
In July of this year, Hassabis said that DeepMind's AI system is helping Google save huge electricity bills. The system reduces power consumption by controlling some of the data center's equipment—manipulating servers and cooling systems. Google also said that the technology does increase power usage efficiency (PUE) by 15%. Google’s electricity consumption in 2014 was 4,402,836 MWh, and a large portion of the electricity was consumed by the data center. According to the US Energy Information Administration, the price of electricity paid by enterprises is about 25~40 US dollars per MWh. No matter which standard is calculated, if the power consumption of the data center is reduced by 10%, it may be Google for several years. Save hundreds of millions of dollars in electricity bills (so that Google’s $600 million in the purchase of DeepMind will soon be “saving†back). Hassabis said: "The AI ​​system controls 120 variables in the data center, including fans, cooling systems and windows." DeepMind may ask Google to add sensors to the data center in the future to further save power through richer information.
L Series Lead Acid Battery,Custom Valve Regulated Battery,12V 38Ah Battery,12V 65Ah Lead Acid Battery
Wolong Electric Group Zhejiang Dengta Power Source Co.,Ltd , https://www.wldtbattery.com