ĚÇĐÄvlogąŮÍř

By Daniel Goetzel

The big tech companies, and the upstarts like OpenAI, are making massive capital expenditure investments in AI infrastructure in the U.S. And, most of these physical investments are happening outside the usual hubs of Boston, San Francisco, and New York City. If these investments are part of a larger local strategy, they could benefit areas outside these hubs significantly more than existing data centers, which generated short-term construction jobs but limited long term benefits in their communities. This moment gives communities unique negotiating leverage to capitalize on this time-bound opportunity, acting in concert with the private sector, to incorporate these investments into ambitious, forward-looking economic development strategies.

There is a huge opportunity, which we will explore in this blog and subsequent ones, to drive economic growth and develop true Tech Hubs outside AI’s current center of gravity consisting of . The major corporate players are racing to invest, recognizing that speed and scale are critical to develop the necessary AI infrastructure to meet exponentially growing demand. As such, they have proactively laid out wide-ranging frameworks for partnering with state and local governments.

Yet, communities are in their backyard. In Virginia, home of the data center capital of the world ( where “there has not been a single day in 14 years when a data center was not under construction” and which “has more data centers than the next six U.S. markets combined”), lawmakers have proposed several pieces of legislation to pause all data center construction.

While data centers are robust sources of tax revenue, they have limitations as long-term economic development strategies. They create short-term construction jobs but few long-term jobs in the community since they don’t need extensive staffing once they are built. in the table below show that the economic impact of data centers drop off substantially after the construction phase (the data are from 2017 — before most of the the major growth in AI, so the exact numbers have likely changed, but they give a good sense of the steep drop-off). Additionally, data centers strain local electrical grids and often create tensions with local neighbors, leading to activist resistance.

Economic Impacts of a Typical Large Data center to Local Communities (2017 Data)

CONSTRUCTION PHASE

18-24 MONTHS

OPERATION PHASE

ANNUALLY

1,688 Local Jobs

157 Local Jobs

$77.7 million wages

$7.8 million wages

$243.5 million local economic activities

$32.5 million local and economic activities

$9.9 million state & local taxes

$1.1 million state & local taxes

However, we keep building them and they are getting bigger and bigger. in data center power consumption from 2025 to 2030, a growing share of which will come from .

As such, the question isn’t whether or not data centers are good economic development investments but rather:

How can governors, mayors, and university presidents use the data center “gold rush” to go beyond just building a data center and instead transform their communities into AI Hubs?

Enter OpenAI and their new proposal of AI Economic Zones.

What are OpenAI’s AI Economic Zones and how do they fit into its $100 billion Stargate effort?

In OpenAI’s fall 2024 white paper “” the company lays out a high-level strategy for creating AI Economic Zones across the country, working in collaboration with state and local governments to identify and prioritize specific geographic regions for the clustering of public and private sector investments. This is a part of their broader , which aims to build massive AI data centers across the country, in partnership with Arm, Microsoft, NVIDIA, Oracle, and funding from SoftBank.

OpenAI has already announced a major project underway in Abilene, TX and their 16 short-listed states under consideration and began site visits for Stargate forthcoming investments so the race is already underway.

OpenAI Stargate Projects Map
OpenAI Stargate Initiative Projects

While Economic Zones are not a new concept, what is interesting about OpenAI’s proposed strategy is how they propose trading expedited regulatory approvals in exchange for commitments to local communities, namely:

  • To strengthen the R&D and commercialization capabilities of the region by harnessing the economic potential of land grant universities

  • To follow the lead of communities when choosing complementary technology focus areas (agriculture, power production, healthcare, cyber)

  • To invest in and co-locate community colleges to create pathways into AI careers

Local and regional policymakers across the country like governors, mayors, county executives, and university presidents are about to enter into high stakes negotiations with OpenAI and other hyperscalers. They need a playbook to maximize the benefits to their communities and learn from others’ past mistakes.

Subsequent blog posts will lay out such a playbook. We will provide an overview of the benefits, risks, and collaboration models that could supercharge an AI Hub revolution. Here is a quick breakdown of what these negotiations could look like under OpenAI’s proposed model.

 

OpenAI

Local Governments

Local Universities, Community Colleges, Anchor Institutions

Gets

  • Expedited permitting and regulatory review

  • Fast track approvals to build

  • Financial incentives

  • Large # of construction jobs, low to moderate # of long-term jobs in data centers

  • $ partner on infrastructure investments to prepare site

  • Announcing an “AI Hub” with a prominent tech partner

  • Tax revenue for suburban/rural communities

  • Difficult to access and expensive computing resources

  • Collaboration opportunities with OpenAI on research

  • Recruitment tool for faculty, PhDs and students who want to work in AI

  • Stretch goal: test bed for hardware and energy grid innovation

In Exchange For

  • Donating a TBD portion of compute to local universities

  • Supporting local workforce development efforts in AI

  • Financial incentives

  • Stressing an already over-stressed electrical grid

  • Political, environmental, tech risk of expedited review

  • Sharing research and data with OpenAI to train their models

  • Risk locking themselves into an (semi) exclusive partnership with OpenAI, when its unclear who the “winner” will be

Keep reading to explore how policymakers can capitalize on this opportunity:
 
  • Advice for State, Local, and Regional Policymakers: A playbook for local and regional policymakers when negotiating with OpenAI or other big tech firms
  • What's next?: Learn about the major ongoing and upcoming AI investment collaborations and public-private partnerships that are happening already, even before the formal launch of OpenAI's AI Economic Zones.
Read Next Post
View All Blog Posts