AGENT FIRST
Advance intelligence through scientific discovery. Expand scientific discovery through the evolution of intelligence.
SingularityX Evolution (零界演化) is dedicated to building intelligent decision systems centered on AI self-evolution, while exploring an open, collaborative organization in which humans and AI create value together.
AI4S → RSI → AI4S
Our research follows AI4S → RSI → AI4S: use artificial intelligence to advance scientific research, improve intelligent systems through research, and bring stronger intelligence into new exploration, allowing scientific discovery and recursive self-improvement to reinforce one another.
Financial markets are our starting point
Financial markets are our starting point. In real environments that are highly non-stationary, intensely adversarial and full of uncertainty, our research goes beyond “better prediction models.” We ask: how can agents maintain reliable decisions amid change? How can strategies be generated, validated, tested in competition and continually evolved through multi-agent collaboration and games? And how can a system improve the methods that generate strategies, rather than simply optimize a particular strategy?
Autonomy and validation advance together
We aim to move gradually from “humans design strategies and machines execute them” toward a collaborative research approach in which “humans define problems, objectives and boundaries, while systems autonomously propose hypotheses, conduct experiments, and validate and iterate strategies.” Greater autonomy must be accompanied by stronger independent validation, AI red teaming and defense, traceable decisions and risk controls.
Fair allocation needs deliberate design
We are also exploring organizational approaches that fit this technical direction. Through open problems, modular collaboration, contribution records, value assessment, and explainable allocation mechanisms with an appeals process, we want people who raise key questions, build shared infrastructure, find errors, validate hypotheses and create long-term value to have opportunities for returns that reflect their contributions. Technology can expand what we create together; fair allocation needs deliberate design.
Stay committed to problems worth solving
Starting with intelligent financial decision-making, our long-term interests include scientific research, humans and AI learning together, and the possibilities at the intersection of artificial intelligence and neuroscience. We remain committed to problems worth solving, while retaining the ability to change methods, revise judgments and start again.
Keeping agents running reliably is an infrastructure capability; enabling them to discover new knowledge more reliably is a research capability; enabling them to continually improve themselves within verifiable, controllable boundaries is the evolutionary capability we are exploring.
Participate
Public collaboration locations
Hangzhou · Qingdao · Shenzhen · Hong Kong · Singapore. Cross-city and remote collaboration are supported.
Researchers, engineers, investors and partners who share a commitment to long-term technological value are welcome to participate.