# 🧠 Distributed Science — The Scientific Process as Multi-Scale Active Inference

**Francesco Balzan, John Campbell, Karl Friston, Maxwell J.D. Ramstead, Daniel Friedman, Axel Constant** (2023) · *OSF*

[![DOI](https://img.shields.io/badge/DOI-10.31219%2Fosf.io%2Fdnw5k-blue)](https://doi.org/10.31219/osf.io/dnw5k)

---
<!-- Schema.org structured data for search engines -->
<!--
{"@context":"https://schema.org","@type":"ScholarlyArticle","headline":"DistributedScience","abstract":"The scientific process plays out in a multi-scale system comprising subsystems, each with their own dynamics. We formalize the scientific process as multi-scale Active Inference, where individual researchers, laboratories, institutions, and the global scientific community each minimize free energy at their respective scales. This framework accounts for hypothesis generation, experimental design, peer review, and knowledge accumulation as nested inference processes. We discuss how this multi-scal","keywords":["distributed science","multi-scale Active Inference","scientific process","Free Energy Principle","meta-science","collective intelligence","cultural evolution","distributed cognition"],"author":{"@type":"Person","name":"Daniel Ari Friedman","url":"https://danielarifriedman.com/"}}
-->


## Abstract

> This paper presents a multi-scale description of the scientific process as a distributed system of evidence-seeking activities, applying the free energy principle to understand science as Bayesian belief updating. It argues that scientific practice is shaped by both human and non-human agents, emphasizing the need for an integrated view that combines modern and non-modern perspectives on knowledge production.

## Keywords

`distributed science` · `multi-scale Active Inference` · `scientific process` · `Free Energy Principle` · `meta-science` · `collective intelligence` · `cultural evolution` · `distributed cognition`

## Key Contributions

- Introduces a multi-scale framework for understanding the scientific process as distributed active inference.
- Integrates modern and non-modern conceptions of science to provide a comprehensive account of knowledge production.
- Proposes a simulation approach for scientific practice that can inform the development of augmented intelligence systems.
- Addresses foundational questions regarding the nature of science and its societal implications.

## Methods

- Applies the free energy principle to model the scientific process as evidence-seeking.
- Utilizes Bayesian belief updating to describe the dynamics of scientific cognition.
- Explores the interaction between individual cognitive functions and collective scientific practices.
- Employs a meta-theoretical approach to reconcile different epistemological views on science.

## 🎯 Consulting & Tutoring

**Available for AI Research Consulting and Tutoring.** [Contact Daniel Ari Friedman, PhD](https://danielarifriedman.com/) for collaboration on Active Inference, Bayesian modeling, and computational biology.

## Citation

```bibtex
@article{2023_DistributedScience,
  author = {Francesco Balzan, John Campbell, Karl Friston, Maxwell J.D. Ramstead, Daniel Friedman, Axel Constant},
  title = {{Distributed Science — The Scientific Process as Multi-Scale Active Inference}},
  journal = {OSF},
  year = {2023},
  doi = {10.31219/osf.io/dnw5k},
}
```

## File Inventory

- `AGENTS.md` (1,837 bytes)
- `2023_DistributedScience.pdf` (591,848 bytes)
- `README.md` (1,488 bytes)
- `SKILL.md` (1,831 bytes)