The Demon in the Machine: How Hidden Webs of Information Are Solving the Mystery of Life
Source: https://press.uchicago.edu/ucp/books/book/chicago/D/bo28628698.html ↗
Davies, a theoretical physicist and astrobiologist, synthesises the most current thinking on the relationship between information and life.
He argues that understanding living systems requires a new concept of information that goes beyond Shannon's mathematical formalism — one that accounts for meaning, context, and causal power.
The book connects Maxwell's demon and thermodynamics to epigenetics, quantum biology, and the search for a definition of life that could guide astrobiology.
Davies engages seriously with the work of Walker, Cronin, and other researchers attempting to formalise the difference between living and non-living matter in informational terms.
Written two decades after Loewenstein's Touchstone of Life, it represents the most updated and accessible treatment of the information-life connection, incorporating developments in systems biology and the physics of information that have emerged since the turn of the century.
Central argument
Davies argues that life cannot be understood through physics and chemistry alone — it requires a fundamentally new concept of information that is not merely syntactic (as in Shannon's formalism) but semantic: information that carries meaning, operates causally, and is context-dependent. His central thesis is that living systems are distinguished from non-living matter by their capacity to store, process, and act on information in ways that close the loop between description and causation — what he calls 'informational closure.' Drawing on Maxwell's demon, thermodynamics, epigenetics, and quantum biology, he contends that this informational layer is a genuine physical phenomenon, not a metaphor, and that formalising it is the key unsolved problem in the definition of life.
Critique
Davies writes with admirable clarity and ambition, but the book's central concept — semantically meaningful, causally efficacious information — remains frustratingly under-formalised. He gestures toward the work of Walker and Cronin as promising formalisations, but stops short of committing to any rigorous framework, which means the core thesis risks being a promissory note rather than a demonstrable claim. A philosophically rigorous reader will note that the gap between 'information that matters to a system' and a physically grounded account of how meaning supervenes on substrate is precisely the hard problem he needs to solve, not merely identify.
Why it matters for product
The distinction Davies draws between syntactic and semantic information maps directly onto a failure mode common in product organisations: teams that optimise for data volume and metric coverage while losing sight of whether any of it is causally connected to user behaviour or business outcomes. His argument that meaning and context are not decorative additions to information but constitutive of its causal power should prompt CPOs to scrutinise whether their measurement systems are tracking signals that actually close the loop — driving decisions — or merely accumulating. More structurally, his framing of living systems as defined by informational closure offers a useful lens for organisational design: autonomous product teams function well precisely when they have the informational autonomy to sense, interpret, and act without the signal degrading through hierarchical translation.