Browse by author
Lookup NU author(s): Dr Xinwei ShiORCiD
Full text for this publication is not currently held within this repository. Alternative links are provided below where available.
The global landscape of AI presents both opportunities and obstacles. Over the decades since the concept of Artificial Intelligence (AI) was initiated, AI has witnessed waves of development, from early attempts to create human-like conversational agents to the recent surge in deep learning and big data. The 21st century has seen remarkable breakthroughs, with applications spanning various industries, including technology, healthcare, and education. Efforts to address AI's impact on society are evident in the development of ethical guidelines and regulations. Countries and regions around the world are working on refining legal frameworks for AI. While research and technology advancements are rapid, the commercialization of AI encounters persistent barriers. Despite significant progress, ethical debates concerning AI's interaction with human society are still heated, and AI applications also grapple with unresolved issues in pertinent scenarios. For the technological innovation problem, large language models, such as those based on the Transformer architecture, highlight the struggle for efficient data utilization and the associated cost of developing advanced algorithms. For the commercialization practice, the burst of the AI investment bubble and discussions about AI potentially replacing traditional labor further complicate the industry's trajectory. As the AI industrial ecosystem evolves, there is a need for coordinated solutions and the development of comprehensive industry standards to propel the AI industry toward sustained and healthy growth. This paper concludes significant hurdles as follows: Firstly, there is a misalignment between AI model demands and industry integration, resulting in a talent gap and high commercialization costs. Secondly, there is a structural imbalance in talent supply and demand, with a shortage of high-quality AI professionals. Lastly, on the supply side of AI models, technical bottlenecks and limitations are hindering broad applications. Additionally, the lack of well-established ethical standards and industry norms globally poses challenges in governance, impacting the acceptance and effective utilization of AI applications. The journey of AI from its conceptualization to the present day reflects a continuous struggle between technological advancements and the complexities of societal integration. The "i7C" framework, which integrates seven key elements: data for computing, computing arithmetic, computing power, computing knowledge, computing scenario, computing talents, and ethics in computing, is proposed to address these obstacles. The "i7C" framework builds a robust AI infrastructure with hardware capabilities such as data for computing, computing arithmetic, and computing power. With computing knowledge and scenarios, it promotes the commercial application of AI technologies. Besides, it systematically cultivates computing talents, fully implements industry application standards, actively responds to ethical concerns in society, and forms a strong cultural bond to drive collaborative and mutually beneficial efforts among all members of the ecosystem. Thus, the AI industry ecosystem achieves sustainable development. To achieve the sustainable development of the AI business ecosystem, core enterprises in the AI industry ecosystem should collaborate with leading partners to collectively build AI infrastructure, drive the commercial application of AI technologies, and together foster a sustainable cultural bond to attract and stimulate more AI ecosystem partners. The AI industry faces challenges in implementation, unclear regulations, and a costly, low-sharing environment. To address this, the “i7C” framework is essential for AI to integrate seamlessly and empower diverse industries.
Author(s): Rong K, Shi X, Lv R
Publication type: Article
Publication status: Published
Journal: 科学学研究 Studies in Science of Science
Year: 2024
Volume: 2023-4074
Acceptance date: 03/01/2024
ISSN (print): 1003-2053
Publisher: Zhongguo Kexuexue Yu ke-ji Zhengce Yanjiuhui
DOI: 10.16192/j.cnki.1003-2053.20240023.001
Altmetrics provided by Altmetric
