Deep brain chain, for the AI ​​era to provide the power and data support

Reporter: pencil boxn

“Every technology company in the future is an artificial intelligence company, which provides multi-dimensional services to artificial intelligence enterprises through blockchain technology, so as to make the computation of computation flexible, greatly reduce the computing cost, and protect the privacy, etc. Is what we want to do and has always insisted on doing. “Deep brain CEO Ho Yong described to me this way. Deep Brain Chain is the first artificial intelligence blockchain landing project in the world dedicated to applying artificial intelligence to various IoT devices and making ordinary devices have the ability to talk, think and reason like the human brain .n
n
Artificial intelligence companies will face the pain pointn
According to the statistics of “Wuzhen Index: Global Artificial Intelligence Development Report 2016” index, there are more than 5,000 newly-established AI companies in the world from 12 to 16 years, with a total amount of more than 20 billion U.S. dollars, including 10% to 30% of the financing are invested in the cost of computing, which does not include large and medium-sized Internet companies to invest in artificial intelligence departments, such a huge investment to promote the development of the industry. Also let us realize that artificial intelligence products want to achieve better product targets, in addition to the algorithm, is the need for massive amounts of data to be trained. But the more data in the same computing resources equal, the more data means longer training longer training. If there are parameters in the training process is not correct, which requires more training to be repeated. Training time is too long, product testing time is too long, is not conducive to the iterative update of enterprise products is extremely unfavorable, so that the loss of product competitiveness of enterprises, is likely to be eliminated in the industry competition failure. Thus the pain points that these enterprises will face will cause many manufacturers have to invest a lot of money to buy GPU, FPGA and other hardware resources. For most small and medium-sized enterprises, this investment of more than 1 million people is a huge burden.n
While making artificial intelligence products, the deep brain team found that many artificial intelligence companies put nearly 10% -40% of the financing amount into the computational power calculation, and the procurement and maintenance of high computational performance hardware became a heavy burden on the enterprise , Restricting the investment in technology research and development.n
In addition, there is a big problem with artificial intelligence data for the entire industry at present, that is, it is technically impossible to separate the ownership and the usage right. Businesses need to buy data from third-party agencies for training, which means there may be a second resale of data, risks of privacy breaches, and resale of data after second sale. For example, a data vendor company originally had 20 customers, in fact, only sold three, but after the last 20 have got the data, so had to raise the price of a single copy of the data.n
The pain points of these two cores can not be solved by technology alone. Is there a good way to completely solve this pain point, so that artificial intelligence companies more quickly promote technological revolution?n
He Yong began to think about this issue, deep brain chain came into being.n
n
Provide low cost, privacy, flexibility and security to artificial intelligence products to the centralized computing platformn
Deep brain chain created the world’s first is the world’s first and only by the blockchain technology-driven artificial intelligence computing platform. The Deep Brain Artificial Intelligence Computing Node is a multi-modal, full-function node (permanent node) that runs on a large server cluster, an idle GPU server compute node for small and midsize businesses, and a personal idle GPU compute node. Deploying these nodes to a deep brain chain platform enables an AI company to reduce hardware costs by 70%.n
In addition, deep brain chains provide data privacy protection. After submitting computing training requirements, enterprises will not come into contact with the data and are fixed in the platform’s algorithm through smart contracts. They can not be changed and effectively avoid possible potential privacy risks for enterprises when applying data. All nodes on the platform are anonymous, each node has joined the anti-copy mechanism, the artificial intelligence company in the distribution of each node computing and training do not know where they come from; On the other hand for the node, because the data Have been encrypted, the nodes do not know what data they are running. Deep brain chain Shanghai Nerve Network nodes can also dynamically adjust the amount of computing based on user products to meet the flexibility of user computing requirements.n
n
Investment by Golden Sands Venture Capital, with the block chain to develop artificial intelligence boundariesn
In September 2017, DeepBrain won the first round of financing of Jinsha River Venture Capital, Gobi Venture Capital and Qianshi with a total investment of 32 million yuan. This is the first project that Zhu Xiaohu shot in the field of artificial intelligence and the only one to obtain the third place of venture capital in Jinsha River Partners Ding Jian, Zhu Xiaohu, Yang Zhiwei and Gobi Venture Partner Jiang Tao blessing of the AI ​​blockchain project.n
n
Depth artificial intelligence, to promote the development of the industryn
Deep brain chain core team are experts in the field of artificial intelligence, CEO He Yong is Chinese Academy of Sciences, East China Normal University Ph.D., who won the Shanghai computer industry innovation that the title of the character, the first field of artificial intelligence entrepreneurs; CTO comfortable UK Artificial Intelligence Ph.D., University of Nottingham, Master of Artificial Intelligence, Bristol University, Beihang University

Leave a Reply

Your email address will not be published. Required fields are marked *