The Hypothesis Explosion Principle

• M. Sturbl

  Author

Artificial intelligence can produce research proposals faster than laboratories can examine them. Whether that difference will widen over the next decade remains an empirical question. This paper proposes the Hypothesis Explosion Principle: between 2027 and 2035, the production of distinct, experimentally testable AI-assisted hypotheses will grow faster than the completion of their first direct experimental tests in selected physical and biological sciences. A registry-based framework separates raw suggestions, qualified hypotheses, completed tests, and independently replicated support. Its primary forecast is a doubling of the hypothesis-to-test flow ratio between 2027–2028 and 2034–2035 in a fixed laboratory panel, with growth in the hypothesis count itself. Nine secondary forecasts examine testing coverage, unresolved claims, experimental costs, screening, and laboratory automation. The framework gives explicit outcomes that would contradict each forecast and distinguishes a rising stock of untested ideas from a harmful research backlog. Existing studies establish the plausibility of large-scale generation and automated experimentation; they do not establish the proposed future divergence. The paper supplies definitions, a stock-and-flow model, and an evaluation protocol rather than new empirical results. Its central claim fails if experimental testing keeps pace with qualified hypothesis production. A decline in testing coverage alone would not show that science has become less productive.

1. M. Sturbl

   Marc Stirbl is an independent researcher and entrepreneur based in Southern California. He is the owner of Air Conditioners “R” Us, a Costa Mesa-area heating, ventilation, and air-conditioning business, reflecting a practical background in mechanical systems, technical problem-solving, and business operations. Public contractor records identify him as the owner of the company.

   Stirbl has also developed an interest in scientific research and forward-looking questions surrounding artificial intelligence and the future of scientific discovery. His work includes The Hypothesis Explosion Principle, a prospective framework examining whether AI-assisted generation of testable scientific hypotheses could begin to outpace the ability of laboratories to experimentally evaluate them between 2027 and 2035. The paper emphasizes falsifiable predictions, experimental throughput, research bottlenecks, and the growing role of automation in science. 

   Marc Stirbl’s work brings together practical technical experience and independent scientific inquiry, with particular interest in how emerging technologies may reshape the process of hypothesis generation, testing, and scientific discovery.

• FULL VERSION (PDF)

• Volume 1

Copyright (c) 2026 Author(s) retain copyright and articles are published under CC BY 4.0.

This work is licensed under a Creative Commons Attribution 4.0 International License.