Stanford HAI's report shows AI’s impact in science, industry, and education.
Stanford University’s Human-Centered Artificial Intelligence Institute (HAI) produced The AI Index 2021 Annual Report. The report authors include Daniel Zhang, Saurabh Mishra, Erik Brynjolfsson, John Etchemendy, Deep Ganguli, Barbara Grosz, Terah Lyons, James Manyika, Juan Carlos Niebles, Michael Sellitto, Yoav Shoham, Jack Clark, and Raymond Perrault.
Organizations that contributed to the report include representatives from arXiv, AI Ethics Lab, Black in AI, Bloomberg Government, Burning Glass Technologies, Computing Research Association, Elsevier, Intento, International Federation of Robotics, Joint Research Center, European Commission, LinkedIn, Liquidnet, McKinsey Global Institute, Microsoft Academic Graph, National Institute of Standards and Technology, Nesta, NetBase Quid, PostEra, Queer in AI, State of AI Report, Women in Machine Learning, and many individual contributors. Supporting partners to the report include McKinsey & Company, Google, OpenAI, Genpact, AI21 labs, and PricewaterhouseCoopers.
Global investment in artificial intelligence in 2020 was USD 67.9 billion, a 40 percent increase from the year before. In 2020, the United States ranked highest in the private investment with more than USD 23.6 billion. Last year China private investment in AI was USD 9.9 billion, less than half the amount of the United States. The report notes that although China trails the U.S. in private investments, it has “strong public investments in AI” with heaving public spending on AI research and development. At a distant third, the United Kingdom invested USD 1.9 billion in AI private investment in 2020.
Private investment in AI for improving human health has surged. Last year AI private investment worldwide increased by 4.5 times that of 2019 with over USD 13.8 billion invested in “Drugs, Cancer, Molecular, Drug Discovery.” At a distant second, USD 4.5 billion was invested in “Autonomous Vehicles, Fleet, Autonomous Driving, Road” globally in 2020, nearly on par with “Students, Courses, Edtech, English Language” with USD 4.1 billion invested.
The top three industries with the highest AI adoption in 2020 were high-tech and telecommunications, automotive and assembly, and financial services, respectively.
Industry is attracting Ph.D. AI talent. In North America, the number of Ph.D.s in AI that went into industry increased from 44 percent in 2010 to 65 percent in 2019.
AI talent diversity is lacking. A majority, 45 percent, of new U.S. resident AI Ph.D. graduates were white, while only 3.2 percent were Hispanic, and 2.4 percent were African American. In 2019, 64.3 percent of new AI Ph.D.s in North America were international students, and 81.8 percent remained in the United States.
The top three industries in the U.S. with AI job postings in 2020 were “Information,” “Professional, Scientific, Tech Services,” and “Agriculture, Forestry, Fishing, Hunting.” The top three countries with the highest AI skill penetration are India, U.S., and China, respectively.
In 2020 China had the highest share of AI journal citations with a 20.7 percent share, followed by the United States with 19.8 percent share, and the European Union with 11 percent share. This was the first time that China surpassed the U.S. in journal citations. In 2019 China surpassed the U.S. in AI conference publications. However, the United States leads in overall citations of AI conference publications with 40.1 percent share in 2020, followed by China with 11.8 percent share, and the European Union with 10.9 percent share.
The top three regions posting AI-related preprint research on arXiv, an online repository of preprint research studies, are North America with 36.3 percent, followed by East Asia and Pacific with 26.5 percent, and Europe and Central Asia with 22.9 percent.
The technology of artificial intelligence has reached a point where synthetically generated images, sound, and text are difficult to distinguish from real-world data. In 2020, progress in facial and speech recognition, image classification, and video analysis made progress as large-scale surveillance technology is becoming less expensive, faster, and increasingly widely deployed.
The report points out the ethical challenges of AI applications. AI transparency, explainability, accountability, and privacy are frequent themes in constructing ethical guidelines and frameworks.
“Though a number of groups are producing a range of qualitative or normative outputs in the AI ethics domain, the field generally lacks benchmarks that can be used to measure or assess the relationship between broader societal discussions about technology development and the development of the technology itself,” the report authors wrote.
We live in remarkable times. Advances in artificial intelligence (AI) machine learning is revolutionizing industries across the board and is rapidly transforming science, education, and technology now, and in the not-so-distant future ahead.
Copyright © 2021 Cami Rosso. All rights reserved.
A version of this article first appeared on Psychology Today.