Nearly half of all AI development jobs are centered in just six U.S. cities, according to a report by American research and education think tank the Brookings Institution.
The report claims that 47% of artificial intelligence-related jobs, especially those tied to generative AI, are primarily found in San Francisco, San Jose, New York, Los Angeles, Boston, and Seattle.
Although the July 2023 report did not give an exact number, Brookings reported that California’s Bay Area—particularly San Francisco and San Jose—led the pack hosting nearly 60% of new generative AI jobs as of May 2023.
Generative AI is trained on vast amounts of data to recognize and produce new content based on user prompts. The top platforms in generative AI include OpenAI’s ChatGPT, Google Bard, and the generative AI art platforms Stable Diffusion and Midjourney.
“If generative AI is to boost prosperity in more places around the country,” the report’s authors wrote, “it will likely require a degree of intentional investment—most notably, from the federal government—in new regions.”
The Bay Area benefits from its proximity to top universities that conduct AI research, including Stanfordand the University of California, Berkeley, as well as companies investing in and developing artificial intelligence—such as Google parent Alphabet, Facebook, Salesforce, and NVIDIA.
A report by the Brookings Institution last March warned of AI being concentrated in “superstar” cities, and a “winner-take-most” dynamic in six digital service industries that provide online services to their clients.
“The increasing geographic concentration of AI raises the question of whether policy can or should be marshaled to counter it and promote more broadly shared benefits,” the report concluded.
Last month, streaming giant Netflix posted new job listings for AI-centered roles. Headquartered in Los Gatos, California, another Bay Area city, Netflix said it would offer up to $900,000 in salary for its AI Product manager position.
Brookings called on Federal and State officials to act on creating policies in the early stages of AI development to broaden the emerging AI map as the technologies grow.
“Federal leadership will be essential to push against excessive AI sector concentration,” the report said.
The organization suggested several things policymakers could do to foster a wider AI development map, including expanding and better distributing research and development (R&D), expanding the reach of the National Artificial Intelligence Research Institute, and prioritizing inclusive datasets and testing tools.
Brookings points to the open-source community as a model for decentralizing AI development, but says that open-source development alone won’t solve the concentration of artificial intelligence jobs. Ultimately, the federal government needs to get involved.
“Open-source distribution and widespread licensing of large language models may also decentralize the industry,” Brookings said. “But the wide distribution of a digital technology does not necessarily entail the wide distribution of its associated innovation, job creation, and executive leadership.”
Earlier this month, a separate report by Goldman Sachs said that AI investments could reach $200 billion globally by 2025.