March 16, 2022 by Tim Hatton
New innovations and uses for artificial intelligence have created new opportunities around the globe—and with those opportunities come new AI jobs to fill.
Today, Stanford University released its 2022 AI Index, an annual study of AI impact and progress developed by the Stanford Institute for Human-Centered Artificial Intelligence (HAI) in partnership with several other business, academic, and government partners.
Emsi Burning Glass is one of those partners. By going through our database of millions of job postings, we were able to show how AI jobs have affected the job market, especially how different regions have been shaped by this growing industry.
Three places stand out: Singapore, California, and Washington, D.C.
Our job postings data comes from more than 45,000 websites, from which we identify millions of job postings. To identify which were AI jobs, we used our Open Skills Taxonomy to generate a list of over 80 skills associated with AI, ranging from decision trees to IBM Watson to unsupervised learning.
If a posting required one or more of those skills, it counted as an AI job. With that information, we were able to compare the number of AI jobs by location, across industries, and over time.
Singapore remains the global hub for AI jobs. Even after a slight decrease in 2021, it still lists over twice as many AI jobs per capita than any other country. The US still maintains the second-highest rate of total AI jobs, seeing a slight uptick after its first decline last year. The UK saw its first decline in its share of AI job postings between 2020 and 2021, which may be a sign of its maturing AI market.
Overall, the number of AI jobs continues to grow steadily around the world, taking up a larger and larger share of the global labor market.
For years, Singapore has been working toward becoming a leader in the AI market. One major focus of those efforts has been AI Singapore, a program that includes training, funding and other support for local AI initiatives. Since 2013, the first year of our analysis, Singapore’s share of AI jobs has grown nearly 1200% (moving from 0.18% of all jobs to 2.33%).
Looking specifically at the US, we see that AI jobs are overwhelmingly concentrated in California. Its 80,238 jobs in 2021 are over twice as many as those in Texas, which had the second-highest total with 34,011. New York, Virginia, and Washington rounded out the top five.
Looking at the proportion of AI jobs to other postings, though, the District of Columbia had far more AI jobs relative to its total than any state. It was followed by Virginia, Washington, Massachusetts, and California.
Industry and Focus
Within our taxonomy of over 17,000 skills, we created a list of 86 skills affiliated with AI work. Those were sorted into seven clusters:
Just as in past years, machine learning and artificial intelligence make up the greatest share of all job postings under the broader AI umbrella. In the US, both categories were up again in 2021 after a decline between 2019-2020. They composed 0.57% and 0.33% of the total postings in the US, respectively. Below them, neural networks (0.15%) and natural language processing (0.13%) had the next highest shares.
Within specific sectors in the US, information required the largest share of AI jobs, at 3.3%, followed by professional, scientific, and technical services, with 2.59%. Among specific sectors, most stayed consistent between 2020-2021, except for agriculture, forestry, fishing, and hunting, which saw a notable decrease. In past years, AI in that sector had been used for mapping and predicting the makeup and structure of resources in forests, oceans, and other natural environments.
The full 2022 AI Index from Stanford’s HAI is available here, which provides further detail about these trends and also explores other areas of AI’s growth, including its impact on education, business, and policy.
For more on the skills required on job postings across industries and regions, visit our newly-redesigned skills page.