The nonprofit sector is all a-twitter about the coming “third wave” of AI-fueled philanthropy, inspired by an article from Nan Ransohoff. The debates range from “this is the biggest thing ever to hit philanthropy,” to “it will just be business as normal” (giving away ~1%) or “it will be a bust.”
I was just reading a summary of all of the debates, thanks to Nick Martin of TechChange, and I was inspired to test out these options.
I am hereby announcing the foundation of the AI Billions Fund.
Other than the fact that I don’t have a billion dollars, I am reasonably well qualified for this.
- I started two for-profit AI companies in Silicon Valley back in the 1980s, both of which were acquired by public companies.
- I’ve run tech for good nonprofits ever since then, demonstrating how AI can benefit much more of humanity.
- My recent book, Technology for Good: How Nonprofit Leaders Are Using Software and Data to Solve Our Most Pressing Social Problems, is the first book to talk about how to start a nonprofit tech company. In it, I feature 60+ nonprofits using tech (and mostly AI) to meet social needs in spite of limited opportunities to get rich.
Based on my experience, and the hundreds of nonprofits I have had the honor to meet and often help, I know exactly how I would spend a few billion. And, now is your chance to bet on the AI Billions Fund!
The Investment Thesis
I’ve done and seen a lot of venture backed for-profit startups in Silicon Valley, many of which have made money. I’ve done and seen many exciting nonprofits using tech to deliver dramatic social impact.
What I haven’t seen is for-profits delivering venture returns by focusing on the problems of the poorest 90% of humanity.
For fairly obvious reasons, profit-maximizing is tough to pull off when your target customers have next to no money. OpenAI’s pivot from being a nonprofit to a for-profit is just one of a long list of stories about why a primary focus on social impact is much easier to deliver via a nonprofit.
If the benefits of AI are really going to flow to most of humankind while preserving their agency and humanity, we’re going to need a lot more investment in tech for good nonprofits. Nonprofits with tech teams, and their eyes on the social impact prize. Not because they are hoping to get obscenely wealthy, but because they choose to work on more important problems while (hopefully) making a good living.
I’ve met these leaders.
- They’ve found product-market fit.
- They have gained the trust of the communities they serve.
- They have gained the knowledge of how the real world works, and how to change it.
What they haven’t gotten?
Enough money to truly scale their impact. Money that allows them to build their teams and technology. And crucially, to build up the critical data assets to effectively apply AI to these pressing social problems, which barely exist today.
The foundation world doesn’t know what to do with these tech nonprofits. There are few foundations that are actually good at assessing technology social enterprises.
So, my investment thesis is that the social return on investments in nonprofit tech enterprises will be some of the top returns imaginable in philanthropy. When you create an AI/software product that can do twice as much good (or five or ten times) as status quo approaches with a given amount of funding, you get the digital equivalent of inventing a drug that cures a disease.
How Would I Spend a Billion
Here is where I would invest a sizable amount of philanthropic capital:
- Invest seven figure grants in tech for good nonprofits who have already found product/user fit at scale. More than a hundred come to mind, and I assure you they can absorb the money.
- Build critically needed core infrastructure to support AI (and invest in expanding nonprofit data capacity required for AI). It is obvious to me that there is a desperate need for a pipeline to take the latest open source LLM models and adapt them to the needs of the social sector. Building versions to run these on commodity cloud infrastructure and locally on commodity PCs both with and without GPUs.
- Invest in large scale datasets relevant to addressing social problems, under responsible data governance and non-commercial control, such as the free Better Deal for Data standard I helped launch this year.
- Invest in talent close to the social problems being solved. There is no reason Africa, Latin America, or Asia need to import U.S. talent at U.S. price points.
There are a handful of great (and new in the last ten years) funders already working on these priorities. There are far from enough of them.
The Patrick J. McGovern Foundation. The Agency Fund. Co-Develop (a $130M fund backed by Nilekani and Gates) is already investing in open source projects in digital public infrastructure, a narrow set of essential baseline capabilities (identity, payments, data exchange).
The AI Billions Fund would likely co-invest with many of these, but also reach great organizations that fall outside the specific objectives of these funders.
Call to Action
If you are a tech leader on the brink of billions (or already have them), get in touch with me. My free one-session karma consults are renowned in the field. You should have a sizable commitment to the AI Billions Fund in your initial philanthropic impact portfolio, run by someone who has actually been a successful serial tech entrepreneur.
Whether or not you decide to invest in the Fund, I’d be delighted to share my list of outstanding tech for good leaders who need your support!
By Jim Fruchterman, an author, leading social entrepreneur, MacArthur Fellow, and founder of Benetech and Tech Matters, nonprofit technology companies based in Silicon Valley.
