What is DCS?
DCS (Causal Structure Evolution Theory) is a cross-scale research framework that traces one causal mechanism across 13.8 billion years — from the early universe through life, brains, minds, civilization, and artificial intelligence.
The core question: 13.8 billion years ago, the universe contained no life, no brains, no language, no civilization, and no AI. Why did it eventually produce beings capable of understanding the past, predicting the future, and actively changing it?
The Core Mechanism
DCS proposes that every major evolutionary transition shares a common causal pattern:
Possibility expansion, dynamical filtering, structural persistence, encapsulation and coarse-graining, causal compression, new macro-level causal efficacy, new hierarchy formation, then new possibilities opened.
In one sentence: evolution is the process by which persistent causal structures become the starting point for the next layer of reality.
Atoms encapsulate lower-level complexity. Molecules become new combinatorial units. Cells organize vast chemical reactions into self-sustaining wholes. Nervous systems use past information to predict futures. Brains simulate multiple outcomes before acting. Humans then reverse the question: to achieve a desired future, what cause should I create today?
Intelligence as Future-Participation
DCS compresses intelligence into a single phrase: intelligence is the capacity for a not-yet-existing future to participate in the present.
You carry an umbrella because you predicted rain. The rain has not happened. But a model of that future has already changed your behavior. This is the core functional diffeai
philosoprence between systems that merely react and systems that predict.
Higher intelligence adds counterfactual comparison: not just what will happen, but if I do A, future X; if I do B, future Y; therefore I should do B. This is where prediction becomes planning, and planning becomes causal intervention.
Why This Matters for AI
DCS places AI on the same causal continuum as biological intelligence. Large language models already predict next outputs from past context — that is already a form of causal prediction. As AI gains long-term memory, tool use, agent capabilities, robotic interfaces, and real-world permissions, the question shifts: is AI merely predicting information, or is it beginning to continuously alter real-world causal paths?
DCS argues that the key variable for AI is not raw intelligence but causal scope — the range of real-world outcomes an AI system can influence. When prediction speed, decision speed, and intervention speed all accelerate simultaneously, the bottleneck becomes whether goals, boundaries, and error-correction systems can keep up.
The Organizational Experiment
Perhaps the most unusual aspect of DCS is how it was produced. It was developed by one person — Wei Rongjie, founder of a Shenzhen one-person company (OPC) called MINDAS — working with AI assistants as literature reviewers, adversarial critics, coding partners, cross-disciplinary translators, and knowledge organizers.
The experiment is about whether AI has changed the minimum organizational unit of scientific research. Can a single researcher, properly augmented, now attempt questions that used to require an entire institute? DCS is offered as one data point in that experiment.
Open Questions
DCS is explicitly presented as a research framework, not a finished answer. Key open questions include: What is the precise mathematical definition of causal persistence? Does coarse-graining actually produce macro variables with higher predictive value? Is there a clear criterion for hierarchical transition? What experimental result would force DCS to admit it is wrong?
Launch Information
DCS Causal Structure Evolution Theory — Online Launch
Date: September 16, 2026
Theme: In Search of the First Principles of Evolution
Website: mindas.me
Contact: contact@mindas.me / +86 18826562299
Preprints: Zenodo DOI 10.5281/zenodo.22709952, Figshare DOI 10.6084/M9.FIGSHARE.33519052
ORCID: https://orcid.org/0009-0002-0501-5570
Note: This research was conducted with AI assistance as part of the project core methodology. Criticism and attempts at falsification are welcome.
Top comments (0)