PDF(23299 KB)
PDF(23299 KB)
PDF(23299 KB)
我国企业人工智能应用现状与挑战
Industrial Application of Artificial Intelligence in China: Current Status and Challenges
深度学习增强了人工智能（AI）算法的通用性，近年来催生了AI产业的快速发展，但实践表明AI技术和算法在产业领域的落地应用依然面临极大困难；企业用上并用好AI、学术界和产业界协同以解决算法落地困难等问题，受到广泛关注。本文着眼我国AI产业的健康可持续发展，从企业AI应用落地的实际案例出发，梳理业界现状、剖析发展挑战、探讨根本原因、提出应对策略。企业AI落地应用的复杂性表现在业务需求、数据、算法、基础设施、配套方案等多个维度，应用成熟度取决于数据的准备程度及治理水平。在国家宏观层面，有必要构建更友好的AI产业生态环境，促进AI全产业链协同发展；以更有力的具体举措支持AI产业的技术研发，特别是全栈AI、AI基础平台及工具体系、AI根技术等，提高我国AI核心技术的自主可控能力；鼓励企业积极实施数字化转型，采用AI技术进行智能化升级，形成AI产业技术研发、企业AI落地创新的强耦合及双向循环。
Deep learning has enhanced the versatility of artificial intelligence (AI) algorithms. In the last decade, the AI industry has been spawned and developing rapidly. However, practice shows that the application of AI technology and algorithms in the industrial field faces huge challenges. Approaches need to be explored for enterprises to properly use AI and for the academia and industry to effectively collaborate to facilitate algorithm application. The study focuses on the sustainable development of China's AI industry, and presents several practical cases of AI application, through which we analyze the current status, challenges, and their root causes pertaining to industrial application of AI and propose corresponding suggestions. The complexity of AI application for enterprises involves multiple dimensions, including business requirements, data, algorithms, infrastructure, and supporting systems. The maturity of AI application depends on the degree of data preparation and the level of data governance. At the national level, a friendly ecology for AI application should be built to promote the coordinated development of the entire industry chain, and specific measures should be taken to support the research and development of AI technologies that focus on full-stack AI, AI basic platform and tool system, and AI root technology, thus to improve the independence of China's AI core technologies. Moreover, enterprises should be encouraged to actively participate in digital transformation and intelligent upgrading using AI technologies, thereby forming a strong coupling and a two-way cycle between research and application of AI technologies.
人工智能 / 企业场景 / 智能解决方案 / 落地应用 / 全栈AI / AI根技术
artificial intelligence (AI) / enterprise scenarios / intelligent solutions / application / full-stack AI / AI root technology
| [1] |
M Turing A. Computing machinery and intelligence [J]. Mind, 1950, 59: 433‒460.
|
| [2] |
LeCun Y, Bengio Y, Hinton G. Deep learning [J]. Nature, 2015, 521(7553): 436‒444.
|
| [3] |
华凌 . 为什么AI很火, 落地却很难 [N]. 科技日报 , 2021-07-26 06.
|
| [4] |
科技情报大数据挖掘与服务平台 . 2011—2020年人工智能发展报告 [R]. 北京 : 科技情报大数据挖掘与服务平台 , 2020 .
|
| [5] |
Executive Office of the President. Maintaining American leadership in artificial intelligence [EB/OL]. (2019-02-14)[2022-04-15]. https://www.federalregister.gov/documents/2019/02/14/2019-02544/maintaining-american-leadership-in-artificial-intelligence.
|
| [6] |
Commission European. Proposal for a regulation laying down harmonised rules on artificial intelligence [EB/OL]. (2021-04-21)[2022-04-15]. https://digital-strategy.ec.europa.eu/en/library/proposal-regulation-laying-down-harmonised-rules-artificial-intelligence.
|
| [7] |
胥会云 . 金桥金融科技综合发展指数: 中国区块链、人工智能专利数已超美国 [EBOL]. 2018-06-15 [ 2022-04-15 ]. https:www.yicai.comnews5432412.html .
|
| [8] |
王坚 . 数据资源与城市大脑 [J]. 秘书工作 , 2020 1 : 75 ‒ 77 .
|
| [9] |
华为技术有限公司 . AI赋能智慧城市白皮书 [R]. 深圳 : 华为技术有限公司 , 2021 .
|
| [10] |
刘合 . 石油勘探开发人工智能应用的展望 [J]. 智能系统学报 , 2021 , 16 6 : 985 .
|
| [11] |
Liu H, Ren Y L, Li X, al et. Rock thin-section analysis and identification based on artificial intelligent technique [J]. Petroleum Science, 2022, 19(4): 1605‒1621.
|
| [12] |
龚仁彬 , 李欣 , 李宁 , 等 . 油气人工智能 [M]. 北京 : 石油工业出版社 , 2021 .
|
| [13] |
华为技术有限公司 . 知识计算白皮书 [R]. 深圳 : 华为技术有限公司 , 2022 .
|
| [14] |
Zhang D, Mishra S, Brynjolfsson E, al et. The AI index 2021 annual report [EB/OL]. (2021-03-15)[2022-04-15]. https://arxiv.org/ftp/arxiv/papers/2103/2103.06312.pdf.
|
| [15] |
Torralba A, Fergus R, T Freeman W. 80 million tiny images: A large data set for nonparametric object and scene recognition [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2008, 30(11): 1958‒1970.
|
| [16] |
Chen T, Kornblith S, Norouzi M, al et. A simple framework for contrastive learning of visual representations [EB/OL]. (2020-05-05)[2022-04-15]. http://proceedings.mlr.press/v119/chen20j/chen20j.pdf.
|
| [17] |
Devlin J, Chang M W, Lee K, al et. Bert: Pre-training of deep bidirectional transformers for language understanding [EB/OL]. (2018-10-11)[2022-04-15]. https://arxiv.org/abs/1810.04805.
|
| [18] |
Radford A, Kim J W, Hallacy C, al et. Learning transferable visual models from natural language supervision [EB/OL]. (2021-02-26)[2022-04-15]. https://arxiv.org/abs/2103.00020.
|
| [19] |
肖立志 . 机器学习数据驱动与机理模型融合及可解释性问题 [J]. 石油物探 , 2022 , 61 2 : 205 ‒ 212 .
|
| [20] |
Tan C Q, Sun F C, Kong T, al et. A survey on deep transfer learning [EB/OL]. (2018-08-06)[2022-04-15]. https://arxiv.org/abs/1808.01974.
|
| [21] |
Yuan X Y, He P, Zhu Q L, al et. Adversarial examples: Attacks and defenses for deep learning [J]. IEEE Transactions on Neural Networks and Learning Systems, 2019, 30(9): 2805‒2824.
|
| [22] |
Zhou B L, Khosla A, Lapedriza A, al et. Learning deep features for discriminative localization [EB/OL]. (2015-11-14)[2022-04-15]. https://arxiv.org/abs/1512.04150v1.
|
| [23] |
Shen D G, Wu G R, Suk H. Deep learning in medical image analysis [J]. Annual Review of Biomedical Engineering, 2017, 19: 221‒248.
|
| [24] |
Hamet P, Tremblay J. Artificial intelligence in medicine [J]. Metabolism, 2017, 69S: 36‒40.
|
| [25] |
Jumper J, Evans R, Pritzel A, al et. Highly accurate protein structure prediction with AlphaFold [J]. Nature, 2021, 596(7873): 583‒589.
|
| [26] |
Raissi M, Perdikaris P, E Karniadakis G. Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations [J]. Journal of Computational Physics, 2019, 378: 686‒707.
|
| [27] |
Wang Q, Mao Z D, Wang B, al et. Knowledge graph embedding: A survey of approaches and applications [J]. IEEE Transactions on Knowledge and Data Engineering, 2017, 29(12): 2724‒2743.
|
| [28] |
He X, Zhao K Y, Chu X W. AutoML: A survey of the state-of-the-art [J]. Knowledge-Based Systems, 2021, 212: 1‒17.
|
| [29] |
柴天佑 . 工业人工智能发展方向 [J]. 自动化学报 , 2020 , 46 10 : 2003 ‒ 2012 .
|
| [30] |
Sculley D, Holt G, Golovin D, al et. Hidden technical debt in machine learning systems [J]. Advances in Neural Information Processing Systems, 2015, 2: 2503‒2511.
|
| [31] |
Oliveira R. Combining first principles modelling and artificial neural networks: A general framework [J]. Computers & Chemical Engineering, 2004, 28(5): 755‒766.
|
| [32] |
埃森哲公司 . 人工智能应用之道 [EBOL]. 2019-08-14 [ 2022-04-15 ]. https:cloud.tencent.comdeveloperarticle1487158 .
|
| [33] |
赵邦六 , 雍学善 , 高建虎 , 等 . 中国石油智能地震处理解释技术进展与发展方向思考 [J]. 中国石油勘探 , 2021 , 26 5 : 12 ‒ 23 .
|
| [34] |
埃森哲公司 . 2021中国企业数字转型指数 [R]. 北京 : 埃森哲公司 , 2021 .
|
| [35] |
匡立春 , 刘合 , 任义丽 , 等 . 人工智能在石油勘探开发领域的应用现状与发展趋势 [J]. 石油勘探与开发 , 2021 , 48 1 : 1 ‒ 11 .
|
| [36] |
董幼鸿 . 上海城市运行"一网统管"的创新和探索 [M]. 上海 : 上海人民出版社 , 2021 .
|
| [37] |
M Monarch R. Human-in-the-loop machine learning: Active learning and annotation for human-centered AI [M]. New York: Simon & Schuster, 2021.
|
| [38] |
肖立志 . 数字化转型推动石油工业绿色低碳可持续发展 [J]. 世界石油工业 , 2022 , 29 4 : 12 ‒ 20 .
|
/
| 〈 |
|
〉 |
