Blockchain circle

One stop hot information platform

About us:

Blockchain circle provides the latest information about blockchain, digital currency, digital wallet, exchange, metauniverse, bitcoin, Ethereum, contract, financial management and so on, and always pays attention to the latest market...

The mutual fund special committee released the "blockchain +ai" Industry Research Report

Time : 23/03/2022 Author : 74vlnj Click : + -
        The national Internet financial security technology expert committee continued to track the development of blockchain technology, conducted in-depth research on blockchain security, blockchain + supply chain and other fields, and launched a series of reports. This report focuses on the hot spot of "blockchain +ai" industry and is jointly launched with Shanghai Zhenlian company in order to become the research basis for the development of the industry. The English abbreviation of artificial intelligence is "Ai", which mainly studies how to make computers do more intelligent work that only human beings could do in the past. The term AI was first proposed at the Dartmouth society in 1956. In 2015, the Illinois group study in the United States showed that AI intelligence at this stage has reached the intelligence level of 4-year-old children.
        With the continuous maturity and application of artificial intelligence technology, the innovation of technical concept around "ai+" is also constantly put forward, among which the technical concept of "blockchain +ai" is particularly prominent. Blockchain is a new application mode of computer technology, such as distributed data storage, point-to-point transmission, consensus mechanism, encryption algorithm, etc. As the underlying technology of bitcoin, it has the characteristics of decentralization, openness, autonomy, difficult to tamper with information, anonymity and so on, which can effectively make up for the problems of data sharing and data security in artificial intelligence applications. Blockchain can provide artificial intelligence with the function of "chain", so that the data and information needed in the "autonomous" operation of artificial intelligence can be reliably recorded and traceable, making AI more credible and safer.
        It can be said that "blockchain +ai" is the cooperation between new technologies. If the two can be organically combined, greater value will be created. From finance, consumption, medical services to government services, the combination of blockchain and artificial intelligence is gradually penetrating all industries and fields. The cooperation between AI and blockchain will solve many problems. While AI provides data analysis and matching, blockchain will provide a more secure and trusted network. AI is known as a strategic technology leading the future. It is one of the core technologies to enhance national competitiveness and maintain national security. It will also become the core driving force of a new round of industrial reform in economic development.
        In China, the development of artificial intelligence is highly valued. On July 8, 2017, the State Council issued the strategic deployment of the "development plan for a new generation of artificial intelligence", which defines the three strategic goals for the development of China's new generation of artificial intelligence: by 2020, the overall technology and application of artificial intelligence will keep pace with the world's advanced level, become an important economic growth point, and comprehensively support the construction of a well-off society; By 2025, the basic theory of artificial intelligence will achieve a major breakthrough, become the main driving force for China's industrial upgrading and economic transformation, and move forward towards the construction of an intelligent society; By 2030, AI theory, technology and application will generally reach the world leading level, become the world's major AI innovation center, and lay the foundation for economic power.
        According to the "statistical report on the development of China's Internet" released by the China Internet Network Information Center (CNNIC) in 2017, the annual number of applications for AI related patents in China reached 30115 in 2016, and the industrial scale exceeded 10 billion yuan. In 2017, the scale of China's AI industry reached 15.21 billion yuan. The industry grew at a growth rate of 40% - 50% every year. It is estimated that it will exceed 30 billion yuan in 2019, As of June 2017, the total number of AI enterprises in China has reached 592, second only to the United States. In September 2017, the Kirin 970 chip launched by Huawei and the a11soc chip launched by Apple have machine learning processing units, laying a solid foundation for AI hardware.
        At present, the artificial intelligence industry has passed through the technological wilderness period, and is in the stage of commercial scenario application formed by the combination of general technology and industry. According to the current classification statistics of Shanghai and Shenzhen stock exchanges, there are 104 listed companies involved in the concept of artificial intelligence, which basically covers the basic layer, technology layer and application layer of artificial intelligence. Compared with the rapid development of artificial intelligence technology that has experienced more than 60 years, blockchain technology has started less than 10 years, and has just experienced three primary stages, namely:. Starting period: from 2009 to 2012, the encrypted digital currency represented by bitcoin made blockchain technology come into the vision of some geeks and emerging technology enthusiasts, and began to form a certain degree of attention and research around the world.
        Embryonic period: from 2013 to 2017, Ethereum introduced smart contracts on the basis of bitcoin's basic technology architecture, which greatly improved the scalability of blockchain, and blockchain technology began to extend to more industries and fields. Development period: in 2018, blockchain technology began to develop iteratively. The development of the industry focused on the construction of a safer technical architecture and the improvement of better basic performance. Blockchain security, blockchain and artificial intelligence began to be valued by the industry, and some applications gradually began to pilot in various industries around the world. At present, the overall development stage of blockchain technology is similar to the initial stage of Internet development, and it still takes time to accumulate before large-scale applications are launched.
        "Blockchain +ai" is a good application combination of mutual empowerment of emerging technologies. The exploration of blockchain technology in the vertical field of artificial intelligence will help accelerate the landing of emerging technologies and continue to improve in the process of practice. At present, most "blockchain +ai" projects are still in the proof of concept stage or early application stage. While AI provides blockchain with a stronger ability to expand scenarios and data analysis, blockchain technology can provide AI with highly reliable raw data to support its continuous "deep learning". With the high development of artificial intelligence in the future, the distributed, transparent and traceable characteristics of blockchain can also be used to ensure that artificial intelligence is always within the scope of human control.
        This puts forward higher requirements for the technological development process of both. Generally speaking, blockchain technology itself is in the early stage, and the combination with artificial intelligence requires continuous iteration to meet the requirements of artificial intelligence for performance and stability. The combination of blockchain and artificial intelligence has the following seven advantages: first, blockchain can improve the data security of artificial intelligence; Second, blockchain can accelerate the accumulation of data, provide more powerful data support for AI, and solve the problem of data supply of AI; Third, blockchain can solve the problem of data privacy during data collection; Fourth, AI can reduce the power consumption of blockchain; Fifth, blockchain makes AI more trustworthy; Sixth, blockchain helps AI shorten training time; Seventh, blockchain helps to create a more open and fair artificial intelligence market.
        The advantages of the combination of the two sides are detailed as follows:. Blockchain can help AI avoid failures caused by data storage problems. Each node in the blockchain stores complete data according to the chain structure, and each storage node is independent and of equal status. The high redundancy of blockchain and distributed data storage can avoid the occurrence of system level risks. Theoretically, unless all nodes are at risk, the data is safe. In addition, considering the "black box" problem of AI diagnosis, clear who established AI, what data to use for training, and who deployed the final agent are the best prevention and control measures to deal with the possible problems of AI.
        Most AI programs currently used are variants of the "deep learning" algorithm. Bad data content will bring corresponding security risks to AI. Blockchain avoids this problem by recording which core algorithms are developed with which group of training data. More broadly, blockchain can record who wrote the original AI algorithm and what data is used to train the algorithm. Some enterprises will collect massive data for their own development, and refuse to share data because of market competition. As a result, these companies have limited access to data and lack of complete data sets to support, resulting in poor quality of AI products.
        Using blockchain technology, we can use the data ledger to purchase and sell some data. Data with strong reliability and high availability will enable enterprises to produce high-quality computer recognition, speech recognition and other data intensive applications. When a large number of similar data are collected for training AI models, the data will be affected by bias or "over fitting". Data samples will not have typical randomness to represent the characteristics of the population. The performance of the model trained with this type of data is much worse than that trained with more different samples. By introducing blockchain technology, different people and companies can provide credible and different data, and more diverse data samples can be obtained to help AI complete "autonomous" decisions.
        The rapid development of artificial intelligence needs to be based on a large amount of data, which inevitably involves the rational use of personal privacy data, such as deriving private privacy information from public databases and deriving it to the information of other relevant personnel, which is beyond the scope of information that most people agree to disclose. The blockchain adopts asymmetric encryption and authorization technology, and the transaction information is open and transparent. However, the account identity information is highly encrypted, and the data can be accessed only with the authorization of the data owner. Even if it is invaded, it is only a small part of the information content, and the complete personal identity information of the user cannot be obtained. Under the operating environment of AI big data, this technology prevents personal privacy from being violated, It is difficult for illegal enterprises to use user data to seek illegitimate interests.
        At the same time, the combination of blockchain and encryption algorithm can separate data ownership and use rights in the process of data sharing, so that data users can use ciphertext for model training and use, and completely eliminate the risk of original data leakage, so as to open up data islands in enterprises and governments. Blockchain projects using POW consensus mechanism need to consume a lot of power resources. Artificial intelligence can learn algorithms to increase the load of data centers, operate computer servers and related cooling systems, optimize cooling, and effectively manage equipment, so as to reduce power consumption. AI can optimize energy consumption, which has been confirmed by Google, Baidu and other companies. In June 2017, Baidu's intelligent building project saved 250000 kwh of power consumption for Baidu within a month. Deepmind, Google's AI laboratory, used AI technology to help Google reduce power consumption by 15%.
        An AI managed blockchain can provide a decentralized identification for AI agents that are independent of the underlying platform of AI. Every major AI agent can register as a generally recognized node, which will provide a solution for AI identification, similar to today's website certificate, to verify website ownership. An AI managed blockchain can also allow each AI agent to write its active conventional hash function into the blockchain classification, so that agents with encryption keys can check it tamper proof. The AI distributed ledger on the blockchain records what AI has done to ensure that the wrong behaviors of AI are found, analyzed and corrected in a timely manner.
        The tamperability of blockchain makes it almost impossible for AI to "hide its trace" and delete criminal activity data. Finally, the consensus mechanism of blockchain can ensure that AI is under control. Through the public record of AI tasks (which must be verified by multiple blockchain nodes), we can ensure that the operation of AI will not exceed the limit. On the premise of using blockchain technology to ensure the authenticity and reliability of training data, the training time of deep learning of an artificial intelligence can be greatly reduced by means of blockchain distributed data storage. For example, an artificial intelligence training can adopt the mode of model parallel or data parallel, and distribute a single model or data on different machines, so as to reduce the training time.
        AI can also remove the synchronization constraint in synchronous data parallelism and adopt asynchronous parallel mode &mdash& mdash; Artificial intelligence does not need to wait for the mutual confirmation of data in each step of information processing, and can directly carry out the next step of operation, so as to further reduce the time of deep learning and training of artificial intelligence. The core value provided by blockchain is "de trust mediation". If you want to create a self-organizing and self-adjusting artificial intelligence network &mdash& mdash; Then distributed accounting technology is the best way. Google, Tencent, IBM, Facebook and other large technology companies have revolutionized distributed computing &mdash& mdash; Computing tasks are dispersed among multiple virtual machines to achieve efficient scalable task processing.
        However, their distributed processing tools are still very centralized, and focus on the unified scheduling of specific tasks by centralized controllers to achieve very specific goals. The smart contract based on blockchain technology will enable the network of "de trust intermediaries". In this trusted network, the two AI systems can interact safely and reliably without any centralized intermediaries. Blockchain can also provide a reputation system for AI, so that each AI can check its reputation before choosing to trade with other AI. In addition, the disintermediation and high transparency of blockchain will encourage these AI developers to share their data and their products without worrying about some cases of favoring competitors or stealing their intellectual property rights, and ensure that all relevant parties receive appropriate remuneration for their work.
        The problems faced by "blockchain +ai" mainly include two aspects: on the one hand, the shortcomings of AI and blockchain itself cannot be effectively solved after combination; On the other hand, the combination of AI and blockchain may cause the original advantages to be destroyed. For example:. At present, some derivative applications of blockchain have certain policy risks all over the world &mdash& mdash; For example, whether the token associated with blockchain technology will be used to stimulate artificial intelligence development or node management in the future, but there is still great uncertainty about how to define the token, both economically and politically. As two cutting-edge emerging technologies, they are not yet fully mature.
        Whether in terms of the current technical indicators of blockchain or the actual landing of artificial intelligence, there are still uncertainties that need to be faced in order to truly combine the two and achieve landing. At present, the main problems of blockchain are capacity expansion, privacy, and computing power. The mainstream public chain is difficult to support the implementation of artificial intelligence on the chain. Data sharing threatens the interests of large enterprises. By weakening the centralization of data, the competitive advantage of large enterprises over small companies is reduced. If anyone can access these data sets and calculations, then anyone has the opportunity to compete with the world's largest company. Removing these barriers from the technology sector will improve society, but attempts to share the market may upset large companies.
        If anyone has the ability to produce the best AI in the world, the market will be shared with many start-ups and small businesses that are competing for part of the market. Companies and government organizations that previously used user data to develop advertising or business strategies will again be forced to obtain their data in a less direct way. Therefore, large companies may oppose data decentralization and may lobby to maintain the status quo of centralized data sets in AI model development. When using the smart contract of "once running, it cannot be stopped", if there is a vulnerability in the contract code that is exploited by hackers, hackers will make profits through the vulnerability of the smart contract, because the transactions and transactions running on the blockchain are not
Previous:The block chain financial integrated service platform in Baoding City, Hebei Province was launched
Next:No more

Related articles:

© 2005-2032 | Blockchain Circle & & All Rights Reserved    Sitemap1 Sitemap2 If there is infringement, please contact us at: