AI Is Here, for Better and for Worse. This Tech Funder Doesn’t Want to Leave It to the Marketplace

sashkin/shutterstock

In late December last year, the Patrick J. McGovern Foundation announced a new round of grants totaling $66.4 million to roughly 150 recipients developing artificial intelligence and other data solutions that advance the public good. 

That’s a substantial chunk of cash, and it’s very much in line with McGovern’s overall funding ethos: The foundation is an avowed pro-technology organization "dedicated to advancing artificial intelligence and data science solutions to create a thriving, equitable, and sustainable future for all." It’s been a key backer of efforts to increase the diversity of the tech sector, for example, and is also a leading funder of nonprofit tech, including AI applications. 

The grants, ranging from $40,000 to $1.5 million, but mostly of several hundred thousand dollars or more, went to organizations in a dozen countries, including the United States. The grant recipients work in several areas where AI and data science are increasingly making inroads — climate, equity and access, education and media among them. Funding also went to organizations aiming to drive policy and research involving AI and data science. McGovern described the grants as a "landmark investment in an ecosystem of partners working to build AI for the public good."

Perhaps most centrally, as AI continues its rapid ascent in the private sector, these are the kinds of topics and applications that for-profit firms would be likely to ignore — what the McGovern Foundation calls "the missing middle" in the AI development ecosystem. 

I connected recently with Vilas Dhar, president of the McGovern Foundation, to expand on his thoughts regarding philanthropy's role in AI as the speedily evolving technology triggers concerns about its impact on society, from jobs to equity to culture. (And not just for actress Scarlett Johansson, who claims that OpenAI, the organization behind ChatGPT, used her voice without permission for its popular generative AI tool.)

According to Dhar, philanthropy has a unique role to play in ensuring that technology doesn't advance simply for the benefit of businesses and wealthier people while exacerbating the inequalities that harm too many people and communities. 

"Our institutions have to look beyond the technology itself and address issues as profound as what it means to have dignity in our work in an automated society, how we limit and shape systems of power away from technological authoritarianism, and how we ensure shared ownership, not merely access, to the benefits of an AI-enabled world," Dhar said.

Dhar, himself a technology optimist, said that AI has great potential as a tool to benefit society at large. He’s not wrong: IP has covered a number of nonprofit efforts to use the technology to advance areas like climate science and medical research and treatment. The McGovern Foundation is also keyed in there: Several grants from its December round went to climate organizations such as Conservation International Foundation, Digital Democracy and Open Earth Foundation, among others. And more than 20 of the grants went to organizations using AI and data science for medical and health applications, including several in the U.S., such as UC San Francisco, and many to international organizations, including Jhpiego Corporation in Nigeria and Khushi Baby in India.

But as Dhar and so many others point out, we do not have to look far to envision how people might — and indeed already do — use AI in deeply disturbing ways that violate privacy, promulgate bias or disseminate harmful falsehoods.

Given the power and pervasiveness of digital technology and now of AI, society can no longer simply leave tech development to Silicon Valley, as we generally have for the last several decades, eagerly buying computers and smartphones, logging onto social media and consuming whatever comes down the digital pipeline. "We need a cultural shift in how civil society and philanthropy interact with technology," Dhar said. "These tools are reshaping our world — we no longer have the luxury of thinking of tech sector activities as independent silos. It’s our responsibility to curate that transformation to further basic human dignity."

Philanthropy and the private sector share some common DNA in that both channel investments of money and expertise, and that figures prominently into McGovern's AI funding and strategy. "Given the nature of the necessary capital investment to build this capacity, philanthropies can play a platform role, aggregating technical skills, best practices and use cases and sharing them across entire ecosystems of actors," Dhar said.

To serve as the "missing middle" in the pro-social AI and data science ecosystem, the McGovern Foundation also builds AI tools to support nonprofit activities at scale and works with AI companies and nonprofits to support the development of industry-wide standards and best practices. "We build products where the 'invisible hand of the market' no longer functions, creating tools that support purpose in the absence of profit," Dhar said.

One challenge for philanthropy, for example, is to catalyze a shift in how existing troves of data — the raw material that AI learns and draws from — are used. "The vast majority of available data sets are collected because of their commercial viability, reflecting decades of investment by private sector actors," Dhar said. Data sets such as those used in mining and natural resource businesses are a good example. Often, the geospatial information systems that map underground extractable resources are highly detailed, while data about drinkable water and other surface-level resources that are critical to people living in local communities above these resources is sparse.

"Philanthropy can step into a new role collecting and stewarding crucial data representing the most vulnerable as a public good, ensuring that AI systems consider and serve everyone," Dhar said. "Philanthropy has a moral responsibility to ensure that those most vulnerable have access to capital, technology and an ecosystem of support. The critical question is whether we can upskill and transform our efforts quickly enough to make relevant contributions in a fast-moving world."

Not least is the need to broaden the "who" in AI and related technology development. That means involving more people from more communities in the earliest phases of envisioning and developing AI products. "Even as we look into transforming representation in data, we must support scholarships and training programs for underrepresented technologists, ensuring that those who build our AI solutions come from varied backgrounds with unique and collective perspectives,” Dhar said.

It is simply not clear at this point what impacts AI will have on society in the coming years and decades. Will it take jobs currently held by humans? Will it create even more new, better jobs? Will it erode privacy? Can it protect privacy? Will it worsen inequality and wealth disparities within and between countries? It may turn out to be a boon in ways we can't yet foresee, and it may drive serious cultural and economic harm in ways both foreseeable and unforeseeable.

In any case, Dhar said, its development can't safely be left to the private sector. "AI holds tremendous potential for good, but I also recognize that the future I envision, where AI furthers human interests and creates better outcomes, is merely possible, not inevitable.”