Strategic Science and Technology Frontiers, author Strategic Science and Technology Frontiers
Strategic Science and Technology Frontier.
Pay attention to global scientific and technological frontiers, strategic technologies, major scientific and technological breakthroughs, etc., track the latest scientific and technological policies, project plans, R&D trends, long-term planning, etc., and provide the latest scientific and technological information. Focus on ICT and cross-cutting frontiers.
On May 23, 2023, the White House announced a series of new initiatives around the use and development of artificial intelligence in the United States and updated the National Artificial Intelligence R&D Strategic Plan. The plan is a re-update of the 2016 and 2019 editions of the National Strategic Plan for AI R&D, reaffirming the previous eight strategic objectives and adjusting and refining the specific priorities of each strategy, while adding a new ninth strategy to emphasize international cooperation. In addition, the report proposes to assess federal agencies' implementation of the National Artificial Intelligence Initiative Act of 2020 (NAIIA) and the National Strategic Plan for AI R&D.
Strategy 1: Long-term investment in fundamental and responsible AI research
Prioritize investing in next-generation AI to drive responsible innovation, serve the public interest, and keep the United States a world leader in AI. This includes advancing foundational AI capabilities (e.g., perception, representation, learning, and reasoning), working to develop AI that is easier to use and more reliable, and evaluating and managing the risks associated with generative AI.
The strategy contains ten priorities: advancing a data-centric approach to knowledge discovery; promoting federated ML; Understand the theoretical capabilities and limitations of AI; Conduct research on scalable general-purpose AI systems; Develop AI systems and simulation simulations in physical and virtual environments; enhancing the perception capabilities of AI systems; developing more powerful and reliable robots; improving hardware to facilitate artificial intelligence; creating artificial intelligence systems that improve hardware; Embrace sustainable AI and computing systems.
Compared with the priorities of the 2019 version of Strategy 1, the 2023 edition revised "conducting research on general-purpose AI systems" to "conducting research on scalable general-purpose AI systems" and made significant adjustments in concept definition and research content; The research content of "human-like artificial intelligence research" and "development of scalable artificial intelligence systems" has been removed; Three new priorities have been added: "Developing AI Systems and Simulation Simulations in Physical and Virtual Environments," "Promoting a Federal Machine Learning Approach," and "Embracing Sustainable AI and Computing Systems."
Strategy 2: Develop effective methods for human-AI collaboration
Deepen understanding of how to create AI systems that effectively complement and augment human capabilities. Open research areas include the attributes and requirements of successful human-intelligence collaborative teams, methods for evaluating the efficiency, effectiveness, and performance of AI collaborative applications, and reducing the risk of harmful consequences from human abuse of AI-enhanced applications.
The strategy contains five priorities: developing the science of human-AI cooperation; Improve collaboration models and performance metrics; cultivate trust in human-intellectual interactions; pursue a deeper understanding of anthropogenic systems; Develop a new paradigm for AI interaction and collaboration. The 2023 edition of Strategy 2 comprehensively promotes human-intellectual cooperation and collaborative research, highlighting the knowledge and understanding of human-intelligent collaborative science to fully understand human-AI cooperation and collaboration and develop new research paradigms to promote human-intellectual interaction.
Strategy 3: Understand and address the ethical, legal and social implications of AI
Develop ways to understand and mitigate the ethical, legal, and social risks posed by AI to ensure that AI systems reflect U.S. values and promote equity. This requires interdisciplinary research, including protecting values through technical processes and designs, as well as advancing AI interpretability and privacy-preserving design and analysis; Develop verifiable metrics and frameworks for accountability, fairness, privacy, and bias.
The strategy consists of four priorities: investing in basic research, including promoting core values through socio-technical system design and studying the ethical, legal and social implications of AI; understand and mitigate the social and ethical risks of AI; using AI to address ethical, legal and social issues; Understand the broader impact of AI. In contrast to the 2019 Strategy 3 priorities, the 2023 edition emphasizes an understanding of the social and moral risks of AI and insights into its broader impacts, building on the design and construction of ethical AI, with the aim of mitigating the negative impacts of AI while harnessing the positive role of AI.
Strategy 4: Ensuring safe AI and AI security
Learn more about designing AI systems that are trustworthy, reliable, and secure. This requires research to improve the ability to test and validate the functionality and accuracy of AI systems, as well as to protect AI systems from cybersecurity and data vulnerabilities.
The strategy contains two priorities: building safe AI and ensuring the security of AI systems. Compared with the five priorities of the 2019 Strategy 4 edition, namely "improving explainability and transparency", "building trust", "enhancing checksum verification", "defending against attacks" and "achieving long-term AI security and value alignment", the 2023 edition of the report condenses two priorities from the perspective of AI two-way security.
Strategy 5: Develop shared public datasets and environments for AI training and testing
Develop and allow access to high-quality datasets and environments, as well as testing and training resources. A broader, more diverse community using the best data and tools for AI research will foster innovation and equitable outcomes.
The strategy contains four priorities: developing and providing accessible datasets to meet the needs of various AI applications; Develop shared large-scale and specialized advanced computing and hardware resources; Respond test resources to commercial and public interests; Develop open source software libraries and toolkits. In contrast to the 2019 Strategy 5 priorities, the 2023 edition adds a new priority of "developing shared large-scale and specialized advanced computing and hardware resources."
Strategy 6: Measure and evaluate AI technologies through standards and benchmarks
The development of a wide range of AI assessment techniques, including technical standards and benchmarks, is essential to guide and facilitate the development of AI systems.
The strategy contains five priorities: developing broad AI standards; establishing benchmarks for AI technologies; increasing the availability of AI test platforms; Involve the AI community in standards and benchmark development; Develop standards for auditing and monitoring AI systems. Compared to the 2019 Strategy 6 priority, the 2023 edition adds a new priority on "Developing Standards for Audit and Monitoring of AI Systems".
Strategy 7: Better understand national AI R&D workforce needs
Increase opportunities for R&D workforce development to strategically develop the U.S. AI workforce. This requires research to improve understanding of the limitations and possibilities of AI and its associated work, as well as the implementation of educational and training measures to facilitate effective interaction with AI systems.
The strategy contains ten priorities: describing and evaluating the AI workforce; Development of AI teaching materials for each learning stage; supporting AI higher education personnel; Training/retraining of the workforce; Explore the impact of diverse and multidisciplinary expertise; Identify and attract the world's best talent; Develop regional AI expertise and take into account the balanced development of vulnerable regions; study options to strengthen the federal government's AI workforce; integrating ethical, legal and social implications into AI education and training; Communicate the priorities of the federal government workforce to external stakeholders. The 2023 edition of Strategy 7 distills out specific priorities.
Strategy 8: Expand public-private partnerships to accelerate AI development
Work with academia, industry, international partners, and other non-federal entities to promote opportunities for sustained investment in responsible AI R&D and the ability to transfer advanced results.
The strategy contains three priorities: maximizing the synergistic benefits of public-private partnerships; expanding partnerships with more diverse stakeholders; Improve, expand and create R&D cooperation mechanisms. The 2023 edition of Strategy 8 distills out specific priorities.
Added Strategy 9: Establish principled and harmonizable approaches to international cooperation in AI research
A 2022 study by the U.S. National Science Council shows that no country is leading the way in all aspects of science and engineering in today's world. The geographical distribution of research results in the field of artificial intelligence is also becoming increasingly dispersed. Therefore, international cooperation in AI R&D needs to be prioritized to address global challenges such as environmental sustainability, healthcare and manufacturing. Strategic international partnerships will help support responsible AI R&D progress and the development and implementation of international AI guidelines and standards, ensuring that the United States remains the central hub of the AI R&D ecosystem.
Strategy 9 is a new strategy added to the 2023 version and reflects the growing importance of international cooperation for AI R&D. The strategy contains four priorities: fostering a global culture and developing global partnerships for the development and use of trusted AI; supporting the development of global AI systems, standards and frameworks; facilitating the international exchange of ideas and expertise; Encourage the development of AI for the benefit of the world.
END