Evolving AI — ProSocial World

Harvard University · October 8–10, 2026

Evolving AI

A gathering of AI researchers and evolutionary scientists to catalyze serious conversation on human–AI coevolution.

The premise

AI as a major evolutionary transition

AI can be seen as a major evolutionary transition, of the same kind as the origin of cells, multicellular organisms, animal societies, and human language and culture. In each of those cases, a new system of inheritance or a new level of individuality emerged.

Evolutionary science has learned a great deal from these transitions, but little of it is being applied to AI. One reason for this gap is that for most of the twentieth century, evolutionary science was confined to the study of genetic evolution. The study of other evolutionary processes, such as cultural evolution and evolutionary algorithms, is much more recent and only slightly predates the AI revolution.

Modern evolutionary science is specific about the conditions under which groups of self-interested parts, from cells to human societies, come to function as cooperative wholes. Those conditions are directly relevant to systems of AI agents, especially superintelligent ones, and the societies that include them.

The gathering

A select group of renowned experts from two communities

AI research

Experts and practitioners of machine learning who appreciate the need for an evolutionary perspective in the technology they are building.

Evolutionary science

Members of the evolutionary science community who have already begun thinking about and working on AI.

Domain-specific topics like security and alignment from AI, or natural selection from evolutionary science, are important but are already discussed elsewhere in the two communities. This gathering is designed to take full advantage of the combination of the two fields, with breakout groups on each topic guided by leaders in that area.

The Topics

Our fundamental focus areas

i

AI as a Major Evolutionary Transition

What can previous major evolutionary transitions teach us about AI, and what evidence would show that a genuinely new system of inheritance or individuality is emerging?

ii

From Adaptive Agents to Adaptive Wholes

What evolutionary mechanisms can transform independently optimizing AI agents into coherent, cooperative systems, and keep their collective goals compatible with human flourishing?

iii

Regulation, Cheating, and Systemic Breakdown

How do adaptive systems detect and suppress components that benefit at the expense of the whole, and what can cancer, immunity, development, and conflict suppression teach us about agentic AI?

iv

Human–AI Coevolution

How will humans and AI systems reshape one another's cognition, culture, values, capabilities, and environments through repeated feedback and cultural transmission?

v

Steering Open-Ended AI Ecosystems

What practical strategies can achieve goals of diversity, innovation, and evolvability while shaping the selection pressures acting on agents, companies, markets, and states toward socially beneficial outcomes?

Evolving AI Travel Fellowships

The first Evolving AI Travel Fellowship cohort

We are pleased to offer a limited number of travel fellowships for students, early career researchers, and builders to attend this gathering. The fellowship covers travel, accommodation, and meals for the three days of the event. Applications are due August 31, 2026.

Applicants will be asked to answer the following question in 500 words or less:

"What do you think the future of human-AI interaction will look like in 5 years? What role do you think evolutionary science has to play in that future?"

Mentorship from renowned experts

Invited attendees include MacArthur Fellows, technology entrepreneurs, authors of best-selling books, and professors from around the world. The gathering is intimate.

Working sessions with peers and experts

Over three days, you will work alongside your peers and our invited experts on the questions at the center of this gathering — a rare chance to think through them with some of the best minds in both fields.

Applying as a group?

Applications can be submitted as either an individual or a small team of 2 or 3 individuals who have a history of working together on research, building, or both. If you are submitting an application as a group, you will each have to submit a separate application and include a brief explanation for why you are applying collectively, a description of your history working together, and what you hope to accomplish through attending as a group.

Begin Application →

Applications close August 31, 2026