As an important driving force for a new round of scientific and technological revolution and industrial transformation, artificial intelligence is considered to be a major opportunity for the future development of mankind, and has a "head goose" effect with a strong spillover and driving force. This "head goose" effect is manifested not only in the vigorous development of the artificial intelligence industry, but also in the fact that it will become the main driving force for my country's industrial upgrading and economic transformation.
Through 2019, artificial intelligence has gathered a bit of trust in disputes and gathered forces for good. Toward 2020, the artificial intelligence industry ecology that returns to rationality will be more optimized. Artificial intelligence technology will evolve from perceptual intelligence to cognitive intelligence. "People-oriented" will become a new direction for artificial intelligence development and achieve a better intelligent world.
01
Rational return, the artificial intelligence industry ecology continues to optimize
"The growth rate of the artificial intelligence industry reached its peak in 2018 and decreased in 2019, but it still maintains a high growth rate, but it has returned to rationality." Deputy Director, Cloud Computing and Big Data Research Institute, China Academy of Information and Communications Technology Wei Kai said.
At present, my country's artificial intelligence industry is showing a good momentum of vigorous development. Artificial intelligence technology is in-depth development, artificial intelligence application scenarios are constantly enriched, and the construction of artificial intelligence open innovation platforms and artificial intelligence innovation development test areas is in full swing.
The "New Generation Artificial Intelligence Development Plan" issued by the State Council proposes that by 2020 the overall artificial intelligence technology and application will be synchronized with the world's advanced level, the core industry scale will exceed 150 billion yuan, and the artificial intelligence core industry scale will exceed 400 billion yuan by 2025. In 2030, the theory, technology and application of artificial intelligence will generally reach the world's leading level, and the scale of core industries will exceed 1 trillion yuan. In 2019, the government work report mentioned "artificial intelligence" for the third time. From the initial "accelerating" and "strengthening" to the current "deepening", with the further promotion of policies and the further maturity of technology, the artificial intelligence industry will be implemented. Significantly speed up.
In the fields of consumption and services, from e-commerce, medical care, education, finance and other industries, we have seen that artificial intelligence has been widely used and profoundly changed the economic and social outlook.
With the advent of the 5G commercial era, in the past year, the connection efficiency of artificial intelligence technology has been further improved, and some key application technologies have reached the world's advanced level; the industrial chain has moved forward hand in hand, and the industrial ecology has been continuously optimized; deep learning, data mining, automatic programming, etc. Industrial application scenarios, to expand stronger general-purpose capabilities.
In 2019, the development of artificial intelligence is returning to rationality, and the return means that there will be a greater leap forward in the future.
02
Cognitive intelligence, the post-deep learning era is here
In recent years, artificial intelligence has made considerable progress in perceptual intelligence, and has even reached or exceeded human standards in many fields, solving the problem of "listening, speaking, and seeing". However, because it relies more on simple statistical fitting, the AI system cannot actively understand the laws and causal relationships behind the development of things. AI still does not have common sense and the ability of causal logic reasoning. This is also the reason why "artificial mental retardation" appears from time to time. . Therefore, the realization of cognitive intelligence is the core of current artificial intelligence research, which is also the core area of future competition.
Cognitive intelligence will draw inspiration from cognitive psychology, brain science, and human social history, and combine cross-domain knowledge graphs, causal reasoning, deep learning and other technologies to establish an effective mechanism for stable acquisition and expression of knowledge, so that knowledge can be used by machines Understanding and application, to achieve a key breakthrough from perceptual intelligence to cognitive intelligence.
The future direction of the artificial intelligence industry depends on the progress of algorithms. In terms of algorithms, there are already excellent models such as deep learning and neural networks. As far as the overall development of domestic artificial intelligence algorithms is concerned, although engineering algorithms have achieved phased breakthroughs, the level of algorithms based on the cognitive level still needs to be improved. .
In Wei Kai's view, there is still room for optimization based on deep learning algorithms. However, we should see that in the future development of new algorithms, two routes are gradually becoming clear: one route is "deep learning + neuroscience + brain science", and the other is "deep learning + knowledge engineering", namely Have structured knowledge and have a high-start learning process.
In other words, as the path of the post-deep learning era continues to be clear, artificial intelligence will evolve from perceptual intelligence to cognitive intelligence, and artificial intelligence algorithms will enter the post-deep learning era.
03
People-oriented, AI also has a "human touch"
Today, artificial intelligence technology is being "embedded" in various industries, and its impact on society is increasing. But the question that follows is: can we completely trust artificial intelligence?
In the hot news of 2019, there are endless negative news related to artificial intelligence, and the characteristics of artificial intelligence "double-edged sword" are vividly demonstrated. Key words such as privacy protection and algorithm security have become unavoidable topics in the development of artificial intelligence.
"The discretionary power of artificial intelligence technology should be constrained by morals and ethics." Wei Kai said that the world is paying attention to whether artificial intelligence should be embedded in ethical requirements and reflecting on how artificial intelligence governance should be carried out. In reflection and practice, the consensus of “people-oriented” gradually converged and became a new direction for the future development of artificial intelligence.
How to turn the "people-oriented" concept into reality? On April 8, 2019, the European Commission issued the "Artificial Intelligence Code of Ethics", which proposed seven elements for achieving credible artificial intelligence. Including ensuring the autonomy, robustness and safety of people using artificial intelligence technology, ensuring privacy and data management, ensuring the transparency of algorithm systems, and requiring artificial intelligence to provide diverse, non-discriminatory and fair services. This is the essence of the core of trusted AI, and it is the embodiment of "technology for good".
my country also pays great attention to the "people" in the development of artificial intelligence. On June 17, 2019, the National New Generation Artificial Intelligence Governance Professional Committee issued the "New Generation Artificial Intelligence Governance Principles-Development of Responsible Artificial Intelligence", which contains the elements of "people-oriented": fairness and justice, inclusive sharing, and respect Privacy and building a community with a shared future for mankind... At the 2019 World Artificial Intelligence Conference, Tencent Chairman and CEO Ma Huateng also stated in his speech that in terms of AI governance, we must lead all-round governance with "technology for good". Ensure that AI is "knowable", "controllable", "available" and "reliable".
"Machine has a core, humans have a heart". An artificial intelligence society with a "human touch" must not only have more efficient productivity, but also have a more colorful spiritual space, a deeper and longer cultural heritage, a more fair and just system guarantee, and a more tolerant and harmonious humanistic care. Keeping the bottom line and rules of technology, letting technology be used where it should be used, and benefiting human society, this is a new topic that must be paid attention to in the development of artificial intelligence to today and even in the future.
Editor: Cao Qian Wang Sihan
Producer: Peng Chao
Excerpt from [People's Post and Telegraph]