Dusoma: AI for Humanity

Dusoma: AI for the Poorest People on Earth

Audio version available here: Dusoma Team Audio

This update and strategy statement is for the group of 70 people I call the Dusoma Team: the people I’ve been working with on AI, Africa, and humanitarian projects.

This message is for all of you—those involved in AI, refugee advocacy, LGBTQIA+ issues, and other related causes. Over the past few years, I’ve worked closely with many refugees, and there has been a particular emphasis on supporting LGBTQIA+ individuals.

We’ve accomplished a lot. We’ve learned even more. Code name Taza —I won’t name individuals here— we have been working side by side. I often say I work for Taza (they/them) is the Principal Investigator (PI) alongside me on our National Science Foundation grant proposal. While we didn’t receive the grant, we now have ideas about how to move forward—especially through better partnerships with the machine learning community and friends in academia. That’s an exciting prospect.

So, where are we now? What have we accomplished? What are our next steps, especially with the United Nations?

What Is Dusoma?

Dusoma is a vision—an artificial intelligence platform designed in response to systems like ChatGPT. I call it the first “big tech humanitarian organization” – I’ll speak more about this in a separate talk, but in short: ChatGPT is a for-profit tool, focused on profitability. No matter how it’s structured, its core mission is to generate profit and enrich the founders and investors.

Dusoma is different. It’s an AI system designed to serve the poorest people on Earth. When we talk about AI “benefiting all of humanity,” we have to acknowledge the reality of the people Kathryn/I have met over the past two years. Tools like ChatGPT do not serve everyone. For example, how does laying off 100 million people benefit humanity? Answer: It doesn’t.

And how is AI helping people in Gorom Camp—people who are sleeping in mud with nothing but a single blanket and are starving to death? Answer: It isn’t.

Dusoma would be designed for someone who is absolutely impoverished —someone who might ask the AI: How can I get food? How do I stop starving? What should I do? Why am I poor? Then it responds with questions. It listens. It helps. It connects people to funding and aid. If we can find a company that pays a few dollars a day to watch advertisements, we can get thousands of legitimate customers connected to that. Ads for water filtration products, ibuprofen, sanitary napkins.

As we all know, AI can only be helpful if it’s trained on the right questions. So first, we’ve worked to define who this “poorest person on Earth” really is.

Defining the Poorest Person on Earth

With Taza and through our work with queer refugees in east Africa, we’ve developed a definition. This person has absolutely no money—literally not even a dollar. When rations are cut in refugee camps, we don’t just hear it in headlines—we see the literal hunger and starvation first-hand. Refugees are told, There is nothing for you, simply because you have two children (you need 3+ for $4 per month. Or there is nothing for you because you’re gay or trans, and you’re at the bottom of the totem pole.

The situation worsened after the U.S. inauguration on January 20th. Overnight, funding disappeared, and people began to starve. I believe we are on the cusp of mass starvation, and unless something changes, millions may die from hunger or disease in Sub-Saharan Africa

So, who is this person? They are:

Starving
Completely without cash or resources
A refugee—stateless, often holding a refugee passport
Living with violence—from the state or from neighbors
Often queer or LGBTQIA+, which exposes them to additional risks
Located in sub-Saharan Africa, particularly in underserved regions
Living in slums, rural poverty, or urban poverty
Possibly facing mental or physical health challenges with no access to care
Administratively erased—such as individuals assigned female at birth who present as male in order to access basic freedoms, like fetching food for a household of women. If it’s discovered that they are menstruating, they could be discovered, raped, beaten and murdered.

This is the person we center in our work. This is who Dusoma is for.

When we think about this work, one of the final things I want to highlight is the issue of violence from neighbors. The person we’re talking about—the poorest person on Earth—is often queer, LGBTQIA+, and in many cases, transgender. So we’re talking about someone who is a trans refugee, starving, and experiencing violence. This is what I call multi-marginalized. There may be a more formal term, but that’s how I’ve come to understand it. That is our product’s persona, the poorest person on earth.

We are different from other American tech organizations because over the last few months, I’ve had the honor of getting to know many people in this category. We’ve been able to provide a small number of them with modest weekly support—around $25 per week. It’s not much, but it’s significantly more than those around them have. That small amount, accompanied by Taza’s case management, has the power to change lives.

We also have local project managers, including one woman who is something of a mother figure to the queer group in the refugee camp. She’s incredible and focuses on helping a group of about 10 queer and trans individuals living in Gorom camp, South Sudan. That project is in progress.

To those 10 of you in that queer group we call “Gorom Pride” —this is a shout-out. You know who you are. I asked someone recently, “How do people in the camp know that your group is “the trans/queer” group?” The answer: You just know. There’s an unspoken recognition. People cluster in communities for safety, but that also means they are more easily targeted.

We receive video and data from these individuals—it’s powerful. They risk their lives to capture this footage. It’s also where we put into practice our ethical data collection. This is the reason we have not released pictures yet – we have to get the content creators to safety first. There are several exceptions – activists who want their story known. We have supported their GoFundMe and helped them document their progress and continued challenges. As we build an AI for humanity, one of our core principles is that data should be a vehicle for prosperity. People should be paid for contributing data, becoming skilled digital workers rather than passive recipients of donations.

Let me introduce someone whose story exemplifies this: Yere, in Rwanda. He has been working with me for over a year on various experiments. Recently, he earned his driver’s license—an achievement that was tremendously expensive, and took persistence and hard work. Right now, we’re working together to fact-check ChatGPT, asking questions like: Why is it easier to lift someone out of poverty than out of Rwanda?

Early on, we rented Yere an Airbnb in Kigali. That experience introduced us to a young professional in the banking industry who became an invaluable mentor. She now helps navigate job hunting and urban life. Yere is now a seasoned traveler and has gained significant digital skills and firsthand experience of city life. Most importantly, he shared with me that this project gave him something precious: hope—the belief that there is a better life. His case study is ongoing as well, and he is one of the people reading this text.

Working with refugees in Africa has been the most meaningful work I’ve ever done.

Now, on to logistics: I’ll be picking up my UN grounds pass on June 16, 2025, and attending the United Nations Open Source Conference. We’re preparing a working prototype for presentation. It will demonstrate, in a technical but accessible way, what it means to “teach the machine.”

Everything ChatGPT knows came from somewhere. Our goal is to be that “somewhere”—to become the people who teach AI what it should know about humanity and rescuing people from poverty. As Taza and I like to say “all you need to lift someone out of poverty is cash and good advice.”

In the short term, our biggest need is fundraising. I’ve hit the limit of what I can contribute personally. The NSF grant would have been a major boost, and while that didn’t come through, I always had multiple paths planned.

Since it’s Pride Month (June 2025), I’m using this opportunity to amplify the story of Dusoma—AI for Humanity—especially within LGBTQIA+ advocacy networks like the Human Rights Alliance and the New Mexico Transgender Resource Group.

One project I’m especially excited about is the development of a new chapter for a newly released gender transition textbook. The current book is written for a U.S. audience. Our version will focus on gender transitioning in Africa as a refugee, and will incorporate the lived experience of starvation, statelessness, and survival alongside the medical and psychological challenges of transitioning.

To the ~10 transgender refugees watching this video: this is for you. As I told Taza, my goal is to turn every participant into a digital worker—earning a fair, livable wage, with transparency and legitimacy. When governments look at these individuals, I want them to see taxpayers, entrepreneurs, and builders of the local economy.

We’ve already connected with someone in Nairobi who runs a guesthouse and is building up the community. I hope he’s watching this video too. His school is for healing trauma and violence through dance. He was educated in the town I am from in Chicago- from the School of the Art Institute. He had me at hello.

So yes, a lot is happening. But again—the biggest hurdle is funding.

I believe the tech angle is powerful. Let’s see what connections emerge from the UN event. And long-term, many of you know my personal goal: healthcare. I hope to return to the issue of HIV and AIDS, especially as PEPFAR funding is being cut. That program was essential in helping children live HIV-free, and its collapse is devastating.

I also have ideas around compassionate clinical trials. For example, I worked with Dushime’s sister, who was 17 at the time, and we began discussing women’s health—menstrual pads, ibuprofen access, and what it would mean to launch a simple clinical trial. Imagine: a trial for something like ibuprofen, lifting an entire refugee camp of 50,000 people out of poverty through a universal basis income. Think of the positive impact on the local economy, if it was done with humanitarian-not-for-profit principles. Again, we can get paid for our labor, but there is not extra money with which to enrich ourselves.

Eventually, I want to stabilize this program, secure funding, and expand our global impact: addressing neglected tropical diseases, water sanitation, healthcare infrastructure, and the technology needed to support them all.

And one final note- as some of you know, at the end of the day, I identify as an artist. I am writing my second novel and Taza and I are working on a documentary.

The fiction is about a smartphone-based “weapon of peace and prosperity.” For example, it has a tracking device buried inside and it is activated if/when the phone is stolen. Taza and I have written a draft of a fictitious adventure movie in which their TV is stole, wakes up, catches the thief and recruits them to the Dusoma mission.

So I go big. I’m honored to lead this team for as long as I am able, and I do everything knowing that I hope this project survives me and lives on – and people like Taza go from partner to leader.

Thank you for watching this team update. Please reach out with any questions. I’m hoping to bring us all together and start speaking with one voice, and we will continue to speak, write, post and fundraise.

Dusoma Team Update – June 6, 2025