Chris Adami, a professor of microbiology at Michigan State University, has an intriguing perspective on life: he sees it as a ***biological information processing system***. According to Adami, evolution is fundamentally a process of acquiring, storing, and utilising information about the environment in the service of survival and replication.[^1] The more complex an organism, the better it is at processing and storing vast amounts of information—because, let’s face it, surviving in an unpredictable world requires a solid information management strategy. Evolution, in its relentless efficiency, favours the persistence of organisms that successfully embed "fitness information"—useful data about their environment—into their very structure. Life isn’t just passively existing; it’s engaged in a constant feedback loop with the environment, a dynamic process in which both organism and surroundings modify and ***encode each other.*** By encoding, we mean that environmental pressures translate into heritable, structural, or behavioural traits. Evolution shapes the body, behaviour, and sensory-motor patterns of organisms, while organisms, through feeding, breathing, reproducing, and altering their surroundings, shape the environment in return. This ongoing dialogue between life and its habitat drives adaptation and survival. At the heart of this process is the three-step mechanism of natural selection: 1. Variation in traits within a population (mutation) 2. Differential survival and reproduction (reproductive advantage) 3. The spread of advantageous traits over generations (heritability) Since this process encodes statistically relevant patterns into organisms, each phenotype[^2] can be said to embed information about its environment. This information forms an implicit, evolved “know-how”—akin to a key fitting a lock without “understanding” its purpose. As we discussed [[Information|earlier]], every physical arrangement of matter contains *some* kind of information. Over evolutionary timescales, organisms extract what is most relevant for their survival from this vast ocean of data and encode it into themselves in two primary ways: their *physical form* (**phenotype**) and their *decision-making processes* (**inference**). Fish scales reduce friction for swimming, bird feathers improve aerodynamics for flight—both are examples of physical adaptations embedding information about environmental constraints. This, to me, is the essence of knowledge: **Information that has been embodied**—meaning embedded into an organism in the service of its persistence. In a strict sense, this means that an organism physical structure *knows* about its environment simply through the way it has been shaped by evolutionary pressures. However, **embodied knowledge** is so much more than just encoding information. The internal arrangement of a living system creates a model of the environment, encoding patterns about conditions the organism has encountered. When organisms use this embodied knowledge to guide behaviour, they are effectively performing embodied **inference** that does not require a brain, but is is encoded in the organism’s gene expression, physical structure, biochemical pathways, and developmental patterns. ![[predictive-model.svg]] Take, for instance, the humble *E. coli* bacterium. It doesn’t have a brain, yet it moves toward areas of higher sugar concentration. How? Its *embodied inference model* tells it that higher sugar concentration means more energy, and more energy means a better shot at survival. Similarly, flowers turn toward the sun because their biological inference system predicts that absorbing more sunlight leads to more energy production through photosynthesis. In both cases, what we call “behaviour” is simply the result of an organism acting on an internal model of the world. In this sense, all behaviour is the product of active inference—an ongoing process of detecting patterns in the environment and adjusting accordingly. This is where *causal power* comes into play: organisms persist not by reacting randomly, but by making decisions based on internal models that have been fine-tuned by evolution. And here’s the kicker—if an organism is essentially a network of inference systems, it needs a way to distinguish between *inside* and *outside*, *self* and *other*. This, ladies and gentlemen, is how the *Self* emerges. ![[self1.svg]] Each inference framework—whether at the level of a single cell or a multicellular organism—is a curated assembly of knowledge about the environment. On a fundamental level, the ***Self*** is nothing more than an *embodied knowledge structure* that has been honed by natural selection. But don’t let the simplicity fool you. This isn’t just a passive repository of facts—it’s an active, constantly updating model of reality, one whose primary job is ensuring continued existence. And here’s where things get really fun: cognition—traditionally seen as a function of brains—is actually happening at the level of cells. Cells **observe** their surroundings, **orient** themselves based on internal and external signals, **decide** on a course of action, and **act** accordingly. This is essentially the ***OODA loop*** (Observe, Orient, Decide, Act) developed by U.S. Air Force Colonel John Boyd.[^3] And while this framework was originally designed to describe fighter pilot dogfights, it turns out that *every* living system, from single cells to complex organisms, operates under the same principle. A cell repairing damaged DNA, a liver detoxifying metabolic waste, a lung oxygenating blood—all are engaged in problem-solving, making real-time decisions based on the information available to them. And here we arrive at the crux of the matter. If life is fundamentally defined by the *embodied knowledge* encoded in its predictive models, and if these models are based on *incomplete information* about the environment, then uncertainty is inevitable. In other words, ***uncertainty is the gap between our models of the world and the reality they attempt to describe***. No organism, no matter how sophisticated, has perfect knowledge of its surroundings. Instead, it operates with approximations, probabilistic inferences, and best guesses. The better these approximations, the higher the chance of persistence. But uncertainty will always remain, because no model can ever fully encapsulate the infinite complexity of the universe. Or, if we want to get a bit more poetic about it: ***uncertainty is the strained relationship between our embodied knowledge and the boundless cosmos it seeks to understand.*** [[Beyond the Selfish Gene|Next page]] <hr> Footnotes: [^1]: Adami, Chris. What is information? https://royalsocietypublishing.org/doi/full/10.1098/rsta.2015.0230# [^2]: Phenotype: In genetics, the phenotype is the set of observable characteristics or traits of an organism. The term covers the organism's morphology (physical form and structure), its developmental processes, its biochemical and physiological properties, its behavior, and the products of behavior. This is distinct from the genotype, an organism's genetic material. [Wikipedia](https://en.wikipedia.org/wiki/Phenotype)[^3]: <a href="https://en.wikipedia.org/wiki/OODA_loop">OODA loop Wikipedia link</a>