What It Takes to Lead in AI

Editors @ The Prompt @ OpenAI Global AffairsOct 29, 2026

What It Takes to Lead in AI

The best response to an uncertain AI future is to preserve our capacity to shape it

Oct 29

Welcome (back) to The Prompt. What’s in this edition:

Leading in AI means preserving the ability to shape what comes next
Spending on midterm TV ads mentioning data centers hits $196 million — and counting
From Philadelphia to Lagos, we’re helping more people turn AI access into know-how

If this was forwarded to you and you find it helpful, make sure you’re signed up for the next issue.

[Perspective] What it takes to lead in AI

One of Silicon Valley’s greatest competitive advantages is its ability to reinvent itself, reorganizing itself around a fast-changing tech frontier. A decade ago, most of the excitement was around apps, social networks, and digital platforms. Hardware, semiconductors, and physical infrastructure seemed fairly ho-hum, even commodified. Today, AI has transformed the landscape, putting chips, computing power, energy, and data centers at the center of the next technological revolution.

Silicon Valley was able to pivot remarkably quickly because it already had nearly all the ingredients in one place: world-class universities, researchers, engineers, chip designers, cloud providers, entrepreneurs, investors, and a culture that rewards experimentation. The next wave of innovation didn’t require building an entirely new ecosystem from scratch. The pieces were already there, even if nobody knew precisely which ones would matter most.

There is an important lesson here for countries thinking about their AI strategies. Competitiveness isn’t just about being good at today’s technology. It’s about preserving the ability to lead in tomorrow’s. That requires strength across the full AI stack: energy and infrastructure, semiconductors, compute, frontier models, applications, and the talent and institutions that connect them.

It’s tempting for countries to pick a narrow role in this ecosystem. Some argue that they don’t need to develop models or build infrastructure because they can simply become world-class AI adopters. But it isn’t easy to innovate with precision. Today’s competitive advantage may be tomorrow’s commodity, while an overlooked capability can suddenly become indispensable. Countries that neglect parts of the stack risk sacrificing the optionality that makes technological reinvention possible. And countries that depend entirely on others for foundational AI capabilities may find their economic and strategic choices constrained.

For America and its allies, this lesson is critical to the development of democratic AI. The countries that develop AI technologies, rather than merely consume them, will have greater influence over how those technologies evolve, the standards they embody, and the values they reflect.

This creates a challenge for AI labs and policymakers to safeguard the extraordinary combination of assets that made America a technological leader in the first place. That doesn’t mean ignoring the difficult questions new technologies create. It means finding ways to manage them without undermining the foundations of future innovation. We need to build data centers that work for local communities, show how AI adoption can help people in the workplace, and ensure powerful AI models are appropriately governed.

But we should be careful not to confuse managing technological change with retreating from it. America’s technological leadership was never the result of predicting exactly where innovation would go. It came from building an ecosystem capable of pursuing many possibilities at once. The best strategy for an uncertain technological future is to preserve the capacity to shape it. – Adam Cohen, Head of Economic Policy, OpenAI

[Data] Spending on ads mentioning data centers climbs to nearly $200M

Less than a week until the US midterms, estimated spending on TV ads mentioning data centers since August 1 has now reached $196 million, up from $140 million we reported last week.

Battleground concentration: Ohio, Pennsylvania, Wisconsin and Michigan account for 53% of that spending.
The affordability connection: Data centers appeared in 14% of all October TV ad airings focused on the cost of living — and 20% of Democratic-aligned airings on that issue.
The partisan gap: Democratic-aligned advertisers have outspent Republican-aligned advertisers by 6-to-1 in ads mentioning data centers this month (through October 27): $119 million versus $21 million.

[Academy] Helping the world put AI to use

Making AI useful takes more than access to the technology. It takes opportunities to learn how to use it.

That’s the idea behind the OpenAI Academy, which recently marked its second anniversary. In two years, more than 4 million people have engaged with its online courses, and more than 250 trainings have reached small-business owners, teachers, nonprofit leaders, veterans, older adults, students, and journalists.

This week, that work stretches from Philadelphia, Tulsa and Sioux Falls, to New York, Indianapolis and Brussels, with Academy programming for nonprofits, policymakers, developers and other communities. In Lagos, Nigeria, the Academy is partnering with Andela and Miva University on a Codex session, while the OpenAI Forum is gathering its community with Paystack, the Lagos-founded payments company serving businesses across Africa.

The Forum is also hosting meetups in Detroit and Chicago, plus a livestream on using Codex and ChatGPT to monitor endangered wildlife without disturbing it.

With the Forum now at 100,000 members, these free events are helping more people build skills, exchange ideas and put AI to use in their own communities.

[Disclosure]

Graphics created by Base Three using ChatGPT.

Did we earn your interest?

Subscribed