We teach what AI cannot do.
Humanitarians AI is a 501(c)(3) bridge education program. We connect international graduates on OPT with real projects, experienced mentors, and a framework for developing the capacities the AI era most urgently requires — and that most education has never taught.
The framework is called Irreducibly Human.
Learn more at irreducibly.xyz
The curriculum you received was built for a world that no longer exists.
Education optimized for fact retrieval, arithmetic speed, and syntactic correctness — the skills that industrial economies required. Machines are now superhuman at all of them. Not faster-than-average. Superhuman, by orders of magnitude, without fatigue, without error.
The curriculum didn't notice. It kept teaching Tier 1. We are training humans to compete on the machine's home turf.
The intelligent response to a cognitive forklift is not to practice lifting heavier objects. It is to learn to operate the forklift — and to develop the judgment to know what needs lifting in the first place.
KEY PROGRAMS & INITIATIVES
Our flagship programs that combine AI technology with social impact to create meaningful change.
FELLOWS PROGRAM
International graduates on OPT develop career-ready capacities through real projects under real stakes. The portfolio is evidence. The judgment, causal reasoning, and collaborative synthesis that the work demands — the things AI cannot supply — are the product. Fellows leave with deployed tools AND the irreducibly human capacities that compound across a career.
Learn moreBOTSPEAK: AI FLUENCY
Our AI literacy program teaches fellows to use AI tools as practitioners — prompt engineering, output evaluation, knowing when to trust a result and when to distrust it. This is Tier 1. It is the entry point. It is necessary and not sufficient. Botspeak is taught as an operational skill: how to conduct the cognitive forklift, not compete with it.
Tier 1 — necessary, not sufficient.
Learn moreLYRICAL LITERACY
Language acquisition is irreducibly embodied and social — rhythm, emotional resonance, communal repetition. These are the mechanisms that make linguistic patterns stick, and no algorithm can replicate them. We use AI to produce content that activates those mechanisms at scale, in partnership with educators and families.
Learn moreAI FOR GOOD
Every AI deployment failure is a Tier 4 failure — the algorithm was running, no one was conducting it. Our fellows build real tools for real organizations AND supply the human judgment layer that makes those tools actually work: problem formulation, output auditing, interpretive judgment. The tools are delivered. The judgment comes with them.
Learn moreDOCUMENTATION INFRASTRUCTURE
Addams: The Learning Documentation System
Named after Jane Addams, who understood that good work without documentation disappears. Addams is the AI-powered documentation system every fellow uses to build their learning record — not their artifact record.
The artifact can now be produced without the learning that should have produced it. Addams documents the learning: the friction, the judgment calls, the moments where the fellow had to supply what the machine could not. That record is the proof of work that matters.
Botspeak is the entry point. Addams documents everything above it.
Learn more about Addams →Addams in action — weekly learning documentation for OPT fellows
Irreducibly Human
What AI Can and Can't Do
A curriculum series, production pipeline, and measurement infrastructure for the cognitive capacities the AI era most urgently requires humans to develop — demonstrated by the method used to build it. The argument evolves through conversation with conversational AI. The curriculum takes shape as the design is debated. What you're looking at is the working document, not the finished product. That's the point.
More on Irreducibly Human →BOTSPEAK
Fluency in AI communication. How to talk to machines — and know when they're talking past you. Prompt engineering, model limitations, hallucination detection, and the rhetorical gap between human intent and machine output.
Explore BotSpeak →CAUSAL REASONING
AI finds correlations. Humans build causal models. Counterfactual thinking, interventionist reasoning, and the kind of "why" questions that statistical models cannot answer.
Explore Causal Reasoning →ETHICAL PLAY
Moral reasoning under uncertainty, value pluralism, and the limits of rule-based ethics. Developing the capacity for ethical judgment that cannot be reduced to optimization or alignment.
Explore Ethical Play →AIMAGINEERING
Generative AI produces outputs. Humans produce meaning. Creative process, aesthetic judgment, conceptual blending, and the difference between novelty and genuine originality.
Explore AIMagineering →EMBODIED TEACHING
Teaching is irreducibly human. Presence, improvisation, emotional attunement, and the embodied skills that make mentorship effective — none of which transfer to a language model.
Explore Embodied Teaching →CONDUCTING AI
The five supervisory capacities: plausibility auditing, problem formulation, tool orchestration, interpretive judgment, and executive integration. The course that teaches you to conduct the orchestra, not play every instrument.
Explore Conducting AI →THE AI SHERPA
A practitioner's guide for experiential learning. The full 18-chapter arc — from design failure diagnosis through the Sherpa infrastructure to domain field guides for co-op, study abroad, clinical, trades, and corporate early career programs.
Explore The AI Sherpa →JOIN OUR MISSION
Your support helps international graduates build the portfolios that launch careers while delivering real AI solutions to communities that need them. Students get jobs. Organizations get tools. Kids learn through song. The public gets free AI education. Mentors shape future talent. 100% of donations fund the programs, mentorship, and project support that make this happen.