UBUNTU VILLAGE · SCIENCE MEETS SPIRIT
The most urgent AI crisis conversation happening right now is missing the people it matters to most.
When Geoffrey Hinton — Nobel Prize-winning computer scientist and one of the architects of modern artificial intelligence — resigned from Google in 2023 to speak freely about AI’s dangers, the world listened. His warnings about superintelligent systems surpassing human control, about AI “black boxes” making consequential decisions no one can audit, about the concentration of AI power in the hands of a few corporations: these concerns are real, and they deserve serious attention.
But while the tech world debates what AI might do in twenty years, Black and brown communities are already living the consequences of AI deployed without accountability, without transparency, and without their voices — today. The existential risk conversation has a blind spot. And that blind spot has a zip code.
What Geoffrey Hinton Said — and Why It Matters
Geoffrey Hinton’s contributions to artificial intelligence are foundational. In 2012, he and his colleagues created AlexNet, an algorithm that transformed AI’s ability to recognize images — a breakthrough so significant it is often called the Big Bang of modern AI. His work on neural networks earned him the 2024 Nobel Prize in Physics. When someone of this stature speaks about AI risk, the scientific community pays attention.
Speaking at the time of his 2023 departure from Google, Hinton warned that AI could surpass human intelligence within the next twenty years, and that the complexity of today’s AI systems — which often contain trillions of parameters — makes it increasingly difficult for anyone, including their creators, to understand or predict how they arrive at their conclusions. These “black box” systems are already being used in healthcare, law enforcement, finance, and criminal justice. The decisions they make affect human lives. And no one can fully audit them.
Hinton has also named the political economy of AI risk directly: powerful Silicon Valley investors have actively lobbied against safety regulations. He cited figures — among them OpenAI CEO Sam Altman — who opposed California’s AI safety bill (SB 1047), which was subsequently vetoed. The message is clear: the people building and profiting from AI are not the same people bearing its risks.
This is important. But it is incomplete. The communities most harmed by AI are not the ones most represented in this debate — and the harms they are experiencing are not hypothetical. They are happening now.
The Harms Already Here: AI and Black and Brown Communities
The AI crisis that Hinton and others warn about as a future threat is, for many communities, a present reality. It does not announce itself as artificial intelligence. It announces itself as a denied loan application, an unwarranted police stop, a misdiagnosis, a rejected job application — with no explanation, no appeal, and no accountable human being on the other side of the decision.
Documented AI Harms in Black and Brown Communities
- Facial recognition misidentification: Studies by the National Institute of Standards and Technology found that facial recognition algorithms produce false positive matches for Black women at rates up to 34 times higher than for white men in some systems. This technology is actively used by law enforcement agencies across the country. Innocent Black men have been arrested based on facial recognition mismatches — including Robert Williams in Detroit and Nijeer Parks in New Jersey.
- Predictive policing: Algorithms used to predict where crimes will occur and who will commit them are trained on historical arrest data — data that already reflects decades of racially biased policing. These systems do not predict crime; they predict where police have previously focused enforcement, concentrating further scrutiny on Black and brown neighborhoods.
- Healthcare AI bias: A landmark study published in Science found that a widely used healthcare algorithm systematically assigned lower risk scores to Black patients than equally sick white patients — resulting in Black patients being less likely to be referred for additional care. The algorithm used healthcare spending as a proxy for health need, without accounting for the documented fact that Black patients have historically received less treatment for the same conditions.
- Hiring and credit discrimination: AI-driven hiring tools trained on historical hiring data reflect historical discrimination. Amazon scrapped an internal AI recruiting tool in 2018 after discovering it systematically downgraded resumes from women — a preview of the deeper reckoning still ahead, the question of whether artificial intelligence will make people jobless before communities ever have a say in how it is deployed. Credit scoring algorithms have been shown to charge Black and Latino borrowers higher rates than equivalent white borrowers.
- Benefits and child welfare systems: AI systems used to determine eligibility for public benefits and to flag families for child welfare investigation have been shown to disproportionately target low-income Black and brown families — including Allegheny County’s widely studied Family Screening Tool — automating poverty penalties at scale.
These are not edge cases or hypothetical scenarios. They are documented, peer-reviewed, and ongoing. The AI black box problem that Hinton warns about is not an abstract future danger for these communities. It is Tuesday. Our companion analysis of algorithmic bias and its hidden cost for communities of color maps this terrain in specific, community-centered detail.
AI Surveillance and the Communities It Targets
In neighborhoods like East Harlem — and in Black and brown communities across the country — AI-powered surveillance is not a distant policy debate. It is infrastructure. Cameras with facial recognition capability have been installed at public housing entrances, including at New York City Housing Authority buildings. License plate readers track movement through neighborhoods. Social media monitoring tools scan the posts of community organizers and activists. Predictive analytics systems flag individuals for increased scrutiny before they have done anything.
This surveillance infrastructure is overwhelmingly concentrated in communities of color, low-income communities, and immigrant communities — the same communities with the least political power to contest it. The same communities whose voices are least represented in the rooms where AI policy is made.
The Algorithmic Justice League, the Surveillance Technology Oversight Project (STOP), and the Electronic Frontier Foundation have documented how AI surveillance disproportionately falls on the same communities that have historically faced the greatest state scrutiny. The pattern is not accidental. AI surveillance does not create racial bias in policing — it scales and automates the racial bias that was already there.
When Hinton warns about the concentration of AI power in the hands of a few corporations or governments, he is describing something that communities of color have already experienced — not as an AI story, but as a policing story, a housing story, a benefits story, a healthcare story. The technology changes. The power structure it serves has remained remarkably consistent.
What Ubuntu Teaches Us About Technology and Humanity
Ubuntu philosophy — Umuntu ngumuntu ngabantu, “a person is a person through other persons” — holds that human beings come into their fullness through relationship, through mutual recognition, through the acknowledgment that our wellbeing is inseparable from the wellbeing of those around us. This is not a sentiment. It is a framework for evaluating every system, every technology, every policy that shapes how we live with one another.
Evaluated through an Ubuntu lens, the current trajectory of AI development fails on its most fundamental premise. AI systems that deny loans without explanation, flag families without accountability, surveil communities without consent, and make healthcare decisions without transparency do not see the people they affect as fully human. They see them as data points. As risk scores. As optimization variables — a strange inversion, given how much energy is spent debating whether machines can truly experience emotions while the humans on the receiving end of their decisions are treated as though they cannot.
Ubuntu philosophy also offers something the current AI safety discourse largely lacks: a communal framework for accountability. When a decision harms someone, the question is not only “what did the algorithm decide?” but “who is responsible to the community, and how is that responsibility discharged?” The answer cannot be “no one, because the system is too complex to audit.” That answer is incompatible with human dignity.
The AI conversation needs Ubuntu. It needs the insistence that every system affecting human lives must be answerable to those lives — not just to its builders, its shareholders, or its most technically sophisticated critics.
What Ethical AI Centered on Community Would Actually Look Like
The communities most harmed by AI are not asking for it to be slowed down or stopped — though some are. Mostly, they are asking for something simpler and more radical: to be included. To have their experiences centered in AI design. To have meaningful recourse when AI harms them. To not be the testing ground for systems they had no say in building.
What Community-Centered AI Would Require
- Meaningful consent and transparency: Communities should know when AI systems are being used to make decisions about them — in housing, healthcare, policing, benefits, and education — and should have the right to contest those decisions with a human being who is accountable.
- Community representation in AI governance: The people most affected by AI deployment must be represented in the bodies that regulate and oversee it — not as token consultees, but as decision-makers with genuine power.
- Algorithmic impact assessments: Before AI systems are deployed in communities of color, they should be required to demonstrate that they do not reproduce or amplify existing racial disparities — the same way environmental regulations require impact assessments before industrial development in communities.
- Moratoriums on demonstrably harmful applications: Facial recognition in public spaces, predictive policing algorithms, and AI-driven benefits eligibility determinations should be suspended until they can be demonstrated to be fair, accurate, and accountable.
- Investment in community-controlled technology: Public funding for AI development should prioritize applications that communities identify as beneficial — health diagnostics, language translation, accessibility tools, educational support — rather than surveillance and control.
None of this is radical. These are the same accountability standards we apply to other systems that affect human lives — medicine, housing, employment law. The question is not whether AI should be held to these standards. The question is why it has been allowed to operate without them for so long — and whose interests that permissiveness has served.
The Future Must Not Be Built Without Us
Ubuntu Village believes that communities are not problems to be solved by technology — they are sources of wisdom, resilience, and vision that any technology claiming to serve humanity must answer to. Our ancestors built systems of mutual care, collective accountability, and shared vision long before algorithms existed. Those traditions are not relics — they are the blueprint for what technology must become if it is to serve life.
The AI conversation is a community health conversation, a racial justice conversation, a human dignity conversation. We intend to keep making that case — and to stand with the communities already living its consequences.
I Am Because We Are. And Together, We Heal.
References + Related Reading
- Fortune — Geoffrey Hinton Wins Nobel Prize in Physics
- CBS News / 60 Minutes — Geoffrey Hinton on AI Dangers
- NIST — Face Recognition Vendor Test: Demographic Differentials
- Science — Dissecting Racial Bias in an Algorithm Used to Manage the Health of Populations
- Algorithmic Justice League — ajl.org
- Ubuntu Village — Why Humanity Feels Colder in 2026
- Ubuntu Village — What to Expect from AI by 2035
- Ubuntu Village — Reparations as Healing: A Public Health Argument — the case for economic repair as community health intervention
- Ubuntu Village — Colonialism Didn’t End: It Became Global Health Policy — how the same power structures show up in health funding
Michele Mitchell is the Founder, President & CEO of Ubuntu Village Inc., a 501(c)(3) nonprofit with programs in Kenya, Uganda, and Nigeria. A writer, advocate, and community strategist working at the intersection of ancestral wisdom, public health, and community power, Michele leads Ubuntu Village’s work to center communities as the protagonists of their own healing. She writes from the conviction that science and spirit are complementary, that healing is relational, and that community is the medicine. Read more about Michele, or connect with her on LinkedIn.
Discover more from Ubuntu Village
Subscribe to get the latest posts sent to your email.