# Life as the Fundamental Physics Problem Adam Frank argues current AI models fail to replicate the self-creating nature of biological life. - **Source:** [Big Think](https://bigthink.com) - **Published:** 2026-11-06 - **Topic:** [science](https://brifing.si/c/science) - **Original article:** - **On Brifing:** https://brifing.si/s/c8105f15-3614-4484-bd98-f04fdb51be69 ## Descartes and Machine Metaphors René Descartes famously dissected living dogs, dismissing their vocalizations as mere mechanical noise. This historical perspective established a reductionist framework that viewed biological organisms as simple machines. Adam Frank, a professor at the University of Rochester, contends that this outdated mechanical metaphor continues to distort our modern understanding of complex living systems. ## Life as Self-Creating Systems Biological life is defined by its capacity for self-creation and internal self-repair. Unlike static mechanical objects, living entities actively maintain their own existence through complex metabolic processes. Frank emphasizes that these organisms possess an inherent agency, allowing them to determine what matters for their survival within their specific environmental niches. ## AI as Artificial Mimicry Current artificial intelligence models are classified by Frank as artificial mimics rather than true intelligence. While these systems process vast datasets to generate outputs, they lack the self-organizing properties found in biological entities. They do not possess the internal drive or the autonomous decision-making capabilities that characterize genuine living, sentient systems. ## Redefining Intelligence in Physics The debate centers on whether intelligence requires a biological substrate or if it can emerge from silicon. Physics traditionally struggles to categorize the unique properties of life, often reducing them to thermodynamic processes. Frank suggests that ignoring the self-creating nature of life leads to a fundamental misunderstanding of what intelligence actually entails. ## The Limits of Computational Models Computational models excel at pattern recognition and data synthesis but remain fundamentally passive. They do not experience the world or prioritize survival, which are core components of biological intelligence. By treating AI as a machine, researchers may be overlooking the essential qualities that differentiate living organisms from sophisticated, non-living algorithmic tools. ## Historical Context of Reductionism The machine metaphor has dominated scientific thought since the seventeenth century, influencing biology, psychology, and now computer science. This reductionist approach seeks to explain complex phenomena by breaking them down into simpler, mechanical parts. However, this method often fails to capture the emergent properties that arise from integrated, self-sustaining biological systems. ## Implications for Future Research If life is indeed the most important problem in physics, then our current trajectory in AI development may be incomplete. Future research must address how self-creation and agency emerge from physical matter. Without incorporating these biological principles, artificial systems will likely remain limited to high-level mimicry rather than achieving true, autonomous intelligence. ## The Future of Astrobiology As an astrobiologist, Frank applies these concepts to the search for extraterrestrial life. Understanding the distinction between a machine and a self-creating system is vital for identifying life on other planets. This philosophical shift challenges scientists to look beyond mere chemical signatures and consider the organizational complexity required for genuine life. ## Sources covering this story - https://bigthink.com/series/full-interview/life-machine-frank/ — --- _Summary written by Brifing from the original reporting. Copyright in the original article belongs to its publisher; read the full piece at the link above._