Alex Ushveridze

Alex Ushveridze

The quest to comprehend the origins of intelligence raises intriguing questions about the evolution of learning abilities in natural systems. Why do living organisms possess an inherent drive to acquire knowledge of the unknown? Is this motivation solely explicable through natural selection, favoring systems capable of learning due to their increased chances of survival? Or do there exist additional, more rapid mechanisms that offer immediate rewards to systems entering the "learning mode" in the "right ways"? This article explores the latter possibility and endeavors to unravel the possible nature of these ways. We propose that learning may have non-biological and non-evolutionary origin. It turns out that key properties of learning can be observed, explained, and accurately reproduced within simple physical models that describe energy accumulation mechanisms in open resonant-type systems with dissipation.

Alex Ushveridze

The fact that accurately predicted information can serve as an energy source paves the way for new approaches to autonomous learning. The energy derived from a sequence of successful predictions can be recycled as an immediate incentive and resource, driving the enhancement of predictive capabilities in AI agents. We propose that, through a series of straightforward meta-architectural adjustments, any unsupervised learning apparatus could achieve complete independence from external energy sources, evolving into a self-sustaining physical system with a strong intrinsic 'drive' for continual learning. This concept, while still purely theoretical, is exemplified through the autoencoder, a quintessential model for unsupervised efficient coding. We use this model to demonstrate how progressive paradigm shifts can profoundly alter our comprehension of learning and intelligence. By reconceptualizing learning as an energy-seeking process, we highlight the potential for achieving true autonomy in learning systems, thereby bridging the gap between algorithmic concepts and physical models of intelligence.

Alex Ushveridze

This paper proposes a novel perspective on learning, positing it as the pursuit of dynamical invariants -- data combinations that remain constant or exhibit minimal change over time as a system evolves. This concept is underpinned by both informational and physical principles, rooted in the inherent properties of these invariants. Firstly, their stability makes them ideal for memorization and integration into associative networks, forming the basis of our knowledge structures. Secondly, the predictability of these stable invariants makes them valuable sources of usable energy, quantifiable as kTln2 per bit of accurately predicted information. This energy can be harnessed to explore new transformations, rendering learning systems energetically autonomous and increasingly effective. Such systems are driven to continuously seek new data invariants as energy sources. The paper further explores several meta-architectures of autonomous, self-propelled learning agents that utilize predictable information patterns as a source of usable energy.

Alex Ushveridze

What is the nature of curiosity? Is there any scientific way to understand the origin of this mysterious force that drives the behavior of even the stupidest naturally intelligent systems and is completely absent in their smartest artificial analogs? Can we build AI systems that could be curious about something, systems that would have an intrinsic motivation to learn? Is such a motivation quantifiable? Is it implementable? I will discuss this problem from the standpoint of physics. The relationship between physics and intelligence is a consequence of the fact that correctly predicted information is nothing but an energy resource, and the process of thinking can be viewed as a process of accumulating and spending this resource through the acts of perception and, respectively, decision making. The natural motivation of any autonomous system to keep this accumulation/spending balance as high as possible allows one to treat the problem of describing the dynamics of thinking processes as a resource optimization problem. Here I will propose and discuss a simple theoretical model of such an autonomous system which I call the Autonomous Turing Machine (ATM). The potential attractiveness of ATM lies in the fact that it is the model of a self-propelled AI for which the only available energy resource is the information itself. For ATM, the problem of optimal thinking, learning, and decision-making becomes conceptually simple and mathematically well tractable. This circumstance makes the ATM an ideal playground for studying the dynamics of intelligent behavior and allows one to quantify many seemingly unquantifiable features of genuine intelligence.