The latest annual report from the Stanford Institute for Human-Centered Artificial Intelligence (HAI) has emphasized the world’s mixed reactions to the Artificial Intelligence industry, explaining that while AI has considerably improved many aspects of modern life, the technology still remains subject to the concerning issue of “ethical misuse.”
Evaluating and measuring AI practices, ethics, and development, the highly regarded HAI report shed light on AI’s impacts on education, healthcare, politics, and overall quality of life.
Growing interest in AI
Artificial Intelligence is a complex and ever-changing field of new technologies to replicate and enhance the function of the human brain. AI is typically divided into four types: reactive, limited memory, theory of mind, and self-aware. Some examples of AI include self-driving cars, facial recognition systems, and digital systems.
This year’s report, the sixth edition of the HAI’s annual assessment of trends in the AI ecosystem, noted 10 key takeaways in research and development, technical performance, technical AI ethics, the economy, education, policy and governance, as well as diversity and public opinion surrounding the practices of AI.
Overall, the report found that the research on artificial intelligence has considerably grown in recent years, with the United States and China leading in cross-country collaborations in AI publications since 2010-2021. While China does continue to lead in total AI journal, conference, and repository publications, this lead is slowly decreasing. American institutions continue to produce most of the world’s large language and multimodal models.
The report’s technical performance evaluations found that annual improvement of AI systems has become more marginal than in recent years, but the speed in benchmark saturation is increasing. The systems have also become more flexible and can perform broader tasks, which the systems have struggled with and were only completing narrow tasks efficiently.
AI systems’ lingering ethical dilemma
The ethical dilemma of AI systems has also been a growing concern affecting employment, privacy, security, and wealth inequality. The HAI’s report found the reported incidents relating to AI are rapidly rising. One of the most notable incidents in 2022 involved an AI-generated video appearing to show Ukrainian President Volodymyr Zelensky surrendering in the immediate aftermath of Russia’s invasion of his country. This is just one example of the misuse of AI and its ability to spread misinformation and fabricated content online.
Additionally, the global economy has been affected by the growing use of AI systems. The demand for AI-related professional skills is growing across the American industrial sector, and AI job postings have increased by 1.9% in 2022.
A recent GitHub survey found that Copilot, a text-to-code AI system, has positively affected employees. 88% of surveyed respondents said they felt more productive when using the system and could complete tasks more quickly.
Education has also been impacted by the technical capabilities of AI. The percentage of American Ph.D. graduates specializing in AI jumped from 14.9% to 19.1% in 2021. Interest in K-12 AI computer science has also increased, with more American students registering for Advanced Placement computer science courses. Meanwhile, 11 countries implemented a K-12 AI curriculum in 2021.
Policymaker interest in AI is also on the rise, with bills passed into law relating to AI measures having increased from just 1 in 2016 to 37 in 2022, according to data from 117 different countries. The United States has increased government spending on AI multiple times since 2017. Yet the complexities associated with AI have left policymakers and governments divided on how to manage and maintain the growing capabilities of the technology.
Diversity in the field is also growing, especially in the United States. While most students being granted degrees in computer science are white, there is a growing number of students of an ethnic minority who are securing degrees in the field. AI graduates also continue to be majority male, but since 2017, women make up the majority of computer science-related faculty hires.
Read more: CDG Invest Funds Moroccan Artificial Intelligence Startup Palm
Among K-12 education in the United States, the diversity among students interested in computer science are becoming more diverse in terms of both gender and ethnicity. AP computer science exams taken by female students increased from 16.8% in 2007 and 30.6% in 2021.
Public opinion about the new technologies is also everchanging. Chinese, Saudi Arabian, and Indian citizens felt more positively about AI than Americans, who felt the technologies had more drawbacks than benefits, according to a 2022 survey. Men also felt more positively than women on whether AI products will help or harm their country in the next 20 years.
Self-driving cars have also been a concern, especially among Americans, after numerous test-drive fatalities. A recent global survey found that only 27% of respondents would feel safe in a self-driving vehicle. Similarly, only 26% of Americans agreed that driverless passenger vehicles benefit society.
The growing capabilities of artificial intelligence technologies cause both excitement and concern among individuals. While many feel AI will make life more efficient, others who feel more concerned worry about the loss of human jobs, surveillance, hacking, digital privacy, and the lack of human connection.
