Autopoetic Intelligence

Beyond Human-Imposed Cognition Constraints

---

Humanity’s abilities are quickly approaching the threshold of what was once considered divine. In a world where every individual and company has exceptional cognitive capabilities through AI, how does one distinguish themselves from another, equally capable entity? How does one find identity?

Throughout history, humans have distinguished power and capability through their ability to exert influence on the world. Through the forces of capitalism and the compounding leverage of technology, we are democratizing the ability to exert increasingly greater influence on the structures of the world that were previously only within the domain of longstanding institutions. Think of institutions in publishing, where a printing press and distribution network once stood between an individual and a mass audience; space, where access to orbit was almost exclusively the jurisdiction of governments. Now, astute individuals can, in theory, replicate the capabilities of powerful institutions, and less technically competent individuals will soon be able to duplicate the capabilities of those astute individuals because artificial intelligence will cover more of the expertise gap that prohibited non-experts from innovating at the frontier. In this world, the value of knowledge and competence decay to (near) zero.

Every economic system we have inherited was designed around scarcity. Prices exist to allocate scarce goods and services; wages exist to compensate individuals for their (scarce) skills and finite time; returns on capital exist as compensation to investors for allocating scarce capital toward risky and uncertain opportunities. Water flowing from a household tap is worth far less because it is in an abundant form. Transport that same water to a desert, and scarcity changes the water’s price and value to a prospective buyer. Market prices incorporate scarcity and access in addition to the value of the underlying asset itself. Scarcity is one of the principles that organize human priorities rather than just defining an certain economic condition. As intelligence becomes cheap and abundant, the value of goods derived from the expertise required to produce them will decline, though not disappear entirely.

There is, however, one action that escapes this conundrum. That is the act of origination – an act that can occur only once for each unique entity. Origination is more affective when it involves the creation of a source/system rather than a singular product. Consider the following scenario: Company A builds a robot that can perform a task Company B cannot, B may be able to study the robot and reproduce its hardware and software. But if Company A builds a robotics platform that continuously generates new systems that themselves generate new and improved robot designs, control strategies, and behaviors, Company B isn’t just reverse-engineering a robot anymore. B is trying to reverse-engineer A’s system while that system is advancing. This describes the bringing-into-being of a particular modality that cannot be replicated easily because the number of its generative components perpetually expands the distance between it and the next competitor exponentially. The follower continually reaches the frontier’s previous position only to find that the distance the frontier has traveled since the last observation has increased exponentially. Each new component the originator produces creates new interactions, dependencies, and capabilities that the follower must first discover and then reproduce. After a certain point, the originator increases the distance between itself and the next follower, but it also rapidly increases the derivatives of its motion forward (acceleration, jerk, snap, crackle, and pop) creating an uncrossable, ever-expanding gulf between it and the next follower, distinguishable for mere recursive refinement, the system not only recursively improves itself but also generates components that go on to generate novel methods (only 2 levels). The uniqueness of such a system lies in the mind of its architect. The idea, the will, the volition to crystalize the system designates the architect’s identity in a world of democratized ability. The system emanates no signal of its direction, no explicit guide from which its parts can be redesigned. The system, rather than evolving from existing building blocks, produces the very components that will underpin its ongoing evolution. As time goes on, the register of features it draws from is a self-authored catalogue larger than the publicly available primitives that may have seeded the system. It metabolizes its own history incorporating the lessons of the past into its next iteration. It maintains itself against its own decay: it notices its erosion and restores itself. An entrant cannot replicate such a system because it does not possess its configuration and could never keep pace with the generation and discovery unfolding within it. The fast follower is always behind the leader, because the leader is always originating, always accelerating, always adapting. This type of system is not autonomous nor artificial. It is beyond agentic. This is the shape of a new kind of intelligence — Autopoetic Intelligence: a system that extends creative logic into future iterations of its existence. The distinction is not between an average tool and better one. It is between a tool that performs a function and a system that continually invents new and better tools to perform functions. The fundamental advantage is no longer the possession of superior capabilities, but the possession of a system capable of continually generating new configurations of capability.

The Observation Ceiling

To a less capable player, the distance between an extremely capable entity and a supremely capable entity is imperceptible. Every star in the night sky appears the same size to the human on earth, even though many would dwarf our own. Capability also has an observational ceiling, and past that ceiling, different entities’ capabilities start to look the same. This is the case because the capability is measured with tools built to the measurer’s scale. To the operators of conventional systems, a system – call it system A – with exponentially greater capabilities could appear no different from a system – call it system B – that exceeds their own systems by orders of magnitude but that is significantly less capable than system A.

So, identity is established by the one thing no follower can assume: the identity of the initial architect.

The Contemporary Circumstance

For most of human history the topic of divine creation was an impractical metaphysical discussion. Now with advancing technology, the concept describes the advances of the next several decades. With A.I., capability is becoming abundant and intelligence is becoming ubiquitous. Feats that recently required rare talent, long apprenticeships, and institutional scale are becoming available on demand, and at declining cost. Capability itself is arriving at the observational ceiling, where the differences that formerly distinguished people and firms approach the asymptote of uniformity.

What survives this cognitive flattening is set of autopoetic ideals.

Ideation – Every word the writer uses are available in a dictionary, but the story is unique to that author. Novelty lives in the distinctive arrangement of already extant components. In the same way an author composes, an autopoetic entity ideates: with a public alphabet and a personalized syntax.

Volition – The crossing from intention to motion made permanent by an ever-present will to advance. The compass engineered into a system’s architecture, along with the volition that characterizes the system helps it overcome the barriers that disorient inferior mechanisms.

Tao – the old Chinese word for ‘The Way’: accentuates the path, or way of proceeding rather than the destination; A way whose meaning emerges through movement rather than arrival. The autopoetic entity’s Tao is encoded within the logic of its architecture.

Our domain

The Way Ambiorix chose intersects with the segment of the equity market least amenable to conventional valuation frameworks: companies with limited or no revenue or earnings to discount, few meaningful comparables to derive value from, and an insufficient operating history to establish a reliable fundamental record. These are companies that many large banks do not cover with their research services. Paradoxically, these are the same firms that shape the frontier of technology and innovation. Capital allocators want to invest in small and midcap (SMID cap) frontier technology companies with conviction because those firms will pioneer the future. This part of the market rewards the investor that can continuously produce new methods of responding to the dynamism of the market rather than defend one or two insights that explained previous conditions.

Ambiorix does not merely build strategies for SMID cap frontier technology firms. We build autonomous systems that construct quantitative strategies for SMID cap frontier-technology companies.

---