"2021 is the year that AI moves from emerging to mature — we're no longer dealing with the exploratory thinking part of scientific research, but rather dealing with things that have a direct positive and negative impact on the real world," said Jack Clarki, co-chair of the Stanford AI Index. ”
On March 16, local time, Stanford University released the "Artificial Intelligence Index Report 2022", pointing out that in 2021, China's academic journals published the largest number of papers, but the number of citations was much lower than that of the United States, ranking third. Cross-border cooperation between the United States and China in the field of artificial intelligence is at an all-time high, growing fivefold since 2010.
The AI Index is an independent program of stanford University's Human-Centered AI Institute (HAI) that publishes an annual report on the AI Index to comprehensively track the latest developments and trends in AI. Since 2017, a group led by Stanford University and formed by a number of experts and professors from MIT, OpenAI, Harvard, McKinsey and other institutions has completed the report.
The report release team summarized the key advances in AI in 2021 into four points:
Private investment in AI has more than doubled since 2020, in part due to larger funding rounds. In 2020, there are four rounds of funding worth $500 million or more; In 2021, there are 15 rounds.
The report notes that global private investment has soared from $46 billion in 2020 to $93.5 billion in 2021. All of this money is going to fewer companies, as the number of newly funded startups has been declining since 2018.
Artificial intelligence has become more affordable and performing at a higher level. Since 2018, the cost of training image classification has been reduced by 63.6% and the training time has been reduced by 94.4%. The median price of robotic arms has also fallen by 46.2% over the past five years.
As the total number of AI publications continues to grow, the United States and China dominate transnational research collaborations on AI. Over the past decade, China and the United States have had the largest number of cross-border collaborations on AI papers.
By 2021, the number of joint papers between the United States and China is 9660, nearly three times that of the second place between the United Kingdom and the United States (3560 papers).
Artificial intelligence patent filings soared, more than 30 times higher than in 2015, with a compound annual growth rate of 76.9%. China ranks first in the world in the number of AI patent applications, accounting for 52% of global patent applications in 2021 and receiving about 6% of grants, roughly the same as in the European Union and the United Kingdom. The United States dominates the number of patents granted, accounting for 40% of the global total.
The 2022 AI Index Report is one of the most comprehensive reports on AI to date. The latest edition of the report includes data from a wide range of academic, private, and nonprofit organizations, as well as more self-collected data and raw analysis than any previous edition, including an expanded chapter on technical performance, a new survey of robotics researchers worldwide, global AI legislative record data from 25 countries, and a new chapter in the in-depth analysis of technical AI ethics metrics.
According to the 2022 Artificial Intelligence Index Report, the field of artificial intelligence is at a critical crossroads. The official website of stanford University's HAI Institute pointed out that in 2021, the globalization and industrialization of artificial intelligence will intensify, while the ethical and regulatory issues of technology will also multiply.
"As AI systems become more powerful, it becomes critical to measure and understand the way they cause harm," said report co-author Helen Ngo.
For the research on AI-related ethical issues and regulation in 2021, the research team summarized the core into five points:
Visual models of large languages and multimodal languages perform well on technical benchmarks, but along with performance improvements, ethical issues are on the rise.
Research on AI equity and transparency has exploded since 2014, with a five-fold increase in publications on related topics over the past four years.
The industry has increased its participation in ai ethics, with publications related to this increasing by 71% at top conferences from 2018 to 2021.
The number of AI-related bills proposed in the U.S. has increased dramatically; lawmakers have introduced 130 laws in 2021, compared to just 1 in 2015. However, the number of bills passed is still very small, with only 2% eventually becoming law in the last six years.
Globally, AI regulation continues to expand. Since 2015, the number of AI-related bills passed by legislatures in 25 countries around the world has increased 18-fold, and the number of mentions of AI in legislative processes has also increased 7.7-fold in the past six years.