Humanity’s fate in WALL-E doesn’t seem far-fetched anymore after Chat-GPT decided what I would have for dinner.

Pixar’s 2008 feature film WALL-E finds humanity on an interstellar cruise away from a polluted earth. The ship’s artificial intelligence, AUTO, pilots the ship and ensures every passenger has “everything they need to be happy,” including plenty of soda, entertainment, and transportation without the pesky responsibility of walking or turning one’s head to talk to their friend.

WALL-E’s insidiously boring dystopia of a world ruled only by robots – humans merely mounds of flesh to be kept fed, docile, and entertained – ignores the potential of artificial intelligence to make our lives better by maximizing, rather than limiting, human potential. After all, WALL-E is not our predestined future – we might instead find ourselves in the cartoon world of the Jetsons.

In such a desired end-state, AI can increase the impact of philanthropists in new and exciting ways, enabling them to:

  1. Streamline the grantmaking process, redirecting funds toward direct impact
  2. Codify the impact of investments in novel, more robust ways

But as philanthropists look to embrace AI, they must ensure that the computers are helping them make the humanity of their missions clear.

In Part I of our AI and Philanthropy series, we focused on three strategic questions education philanthropists must ask to assess their approach and theory of change in the AI era, focusing on clear goals, heightened speed, and greater flexibility.

In Part II of our series, we focused on questions funders should seek to answer when evaluating their portfolio strategy as artificial intelligence changes the landscape not only of their grantees, but also of those grantees’ beneficiaries. We suggested that the advent of widespread use of AI in education and workforce training and development would bring about a change in how philanthropists support and push their portfolio of grantees. We now turn our attention to how AI can improve the grantmaking process. Specifically, we consider three questions philanthropists should explore as they look to the use of AI:

  1. How will AI change the grant writing process?
  2. How could AI change the grant evaluation process?
  3. In what ways can AI change how impact is measured?

Streamlining the grant-writing process

One of the primary advantages of utilizing AI in philanthropy is the streamlining of the application process. By employing Large Language AI models, such as Chat-GPT, non-profits and grant-seekers can:

  1. Make their grant-writing process more efficient
  2. Enable themselves to write more compelling grant applications
  3. Reduce the financial burden associated with writing grants
  4. Allocate more time and resources to their core missions

These changes will drive impact throughout each non-profit’s respective ecosystem. Non-profits whose impact has been hindered by a lack of skill in grant writing may be able to amplify their impact in ways they weren’t able to in the past.

New opportunities for philanthropic improvement

Imagine a literacy non-profit dedicated to improving reading scores that struggles to effectively convey its methods. The organization consistently improves scores by one standard deviation but is still growing and needs a team member who knows how to effectively write grant applications. In a world where AI can enable them to articulate their approach and capture their performance data, their impact might be outsized, as funding decisions tend to favor organizations that can deftly translate their work into an impact narrative. This nonprofit could, for example, feed its financial statements and programmatic results into an AI that would translate those results into the format needed by funders and could also rapidly draw insights and conclusions from those data to present not only how funding is used, but the impact of those dollars. The speed, efficiency, quantity, and effectiveness of their grant writing can be vastly improved, allowing organization to focus more on their core capabilities than on seeking funding.

The humanity of this organization’s work is not lost to AI. Rather, by using technology to speed up and improve their grant writing, education nonprofits can focus precisely on those things that improve human lives.

Reducing resource expenditure for foundations

AI can also benefit philanthropists by outsourcing the labor-intensive task of processing grant applications, allowing foundations to reallocate their resources more efficiently. So, as more philanthropic dollars are directed toward more tangible goals and missions, the sectors in which those philanthropists work will benefit.

Imagine again our literacy nonprofit. It has gathered evidence of its effectiveness, and using an AI model has clearly conveyed its impact. Reviewing hundreds of such applications – the quantity of which will surely increase as AI makes the grant writing process easier – will further tax program officers. A well-trained AI could be deployed to scan all applications for the information funders seek to see which organizations best fit their theory of change. Moreover, such a model would be able to more easily translate applications into board-ready presentations.

Grantmaking, of course, is not a mechanical process. Program officers get to intimately know their grantees and learn about their work in a human way. AI, then, does not obviate those relationships but rather, because it speeds up the grant writing and review process, allows program officers to spend more time seeing and learning about their grantees.

Finding new ways to measure impact

Finally, and perhaps most exciting, AI can vastly improve the ways in which we measure and report impact. In addition to large language models, philanthropists may employ big data models to evaluate impact. Education lacks the ease of measurement of some other sectors. There isn’t a kiloton of carbon, an acre of rainforest, or the number of meals served for education organizations to quickly point to.

Philanthropists can feed AI data models vast amounts of data and train them to pick out and analyze any number of outcomes and impact results. They can compare reading scores with attendance and dozens of other metrics to create new holistic measurements of impact we hadn’t conceived of before. Big data models could revolutionize standards of evidence for philanthropists.

Going on autopilot

The increased acceptance and utilization of AI in philanthropy might bring about a revolution, but if philanthropists aren’t diligent, it might not be one that we want to come about.

If grantees over-rely on AI to help them articulate core pieces of an organization’s impact that are lost in translation coming from the voice of something that has yet to see or feel the impact of an organization’s work personally.

As AI is introduced to the grant-evaluation processes, biases around race, gender, and religion present in past grants that taught AI what to look for might be perpetuated. They might seep their way into the AI’s diagnostic protocol, like the AI that learned to detect the rulers present in images as skin cancer as malignant.

Moreover, although big-data models might work magnificently to predict and measure organizations’ impact, what is going on inside these models might be a black box or biased and confused by its input data.

Keeping one hand on the wheel

The key to navigating the confusing world of AI in philanthropy in the future will be the same as navigating the world of philanthropy has always been: philanthropists who are diligent in their pursuit of great change.

As we learned in Pixar’s WALL-E, if we can reconcile AI’s role alongside humanity without taking our hands off the steering wheel, we’ll be able to accomplish things we thought were otherwise impossible.

We are always happy to discuss as you and your organization think through these challenges and opportunities. Reach out to have a conversation. If you missed them, please be sure to check out part one and part two of our AI series to read about how AI will disrupt philanthropy and your philanthropic portfolio.