paint-brush
LLMs | Sentience: What Does Consciousness Do and How?by@step
118 reads

LLMs | Sentience: What Does Consciousness Do and How?

by stephenAugust 16th, 2023
Read on Terminal Reader
Read this story w/o Javascript
tldt arrow

Too Long; Didn't Read

The summary of these observations is that consciousness helps to know. Knowing is the purpose of consciousness, or knowing is what consciousness does. All the descriptions of sentience or consciousness are knowing outcomes. Self-awareness is a known output of being or the self. Navigation, communication, evaluations, decisions, experiences are all known.
featured image - LLMs | Sentience: What Does Consciousness Do and How?
stephen HackerNoon profile picture

There is a new Correspondence in Nature, Consciousness: unicellular organisms know the secret, where the authors stated that:


“All life is sentient. Both life and sentience involve self-awareness, evaluation of perceived information and mutually reactive sensory and perceptual functions. For our research into the cellular foundations of consciousness, we found it most productive to start with the simplest prokaryotic species.


The data indicate that unicellular organisms are highly social, display associative learning (grasping, for example, navigational routes and simple patterns) and form stable memories. They also make decisions, evaluate events, communicate within and between colonies and, fascinatingly, show a form of altruism.”

Consciousness Helps to Know

The summary of these observations is that consciousness helps to know. Knowing is the purpose of consciousness, or knowing is what consciousness does. All the descriptions of sentience or consciousness are knowing outcomes.


Self-awareness is a known output of being or the self. Navigation, communication, evaluations, decisions, and experiences are all known.


Whatever is in a habitat that is recognized by an organism is a knowing process that could be other organisms or the components of the habitat. The construction of knowing varies across species, with humans having the most properties.


Theoretically, the human mind is the collection of all the electrical and chemical impulses of nerve cells, with their features and their interactions. Consciousness is produced when impulses interact.


It is the specifications of these interactions with their features that decide knowing — or consciousness.


Across the brain, with nerve cells and impulses, knowing processes are ubiquitous, though those defined for consciousness are associated with the thalamocortical pathway.


Conceptually, a significant thing about consciousness, especially in the cerebral cortex, is a feature called drift or stairs of sets of chemical impulses, where rationing or fills occur.


Synaptic clefts are established to include vesicles and receptors.


It is postulated that in sets, they represent rationing stations that distinguish a taste from a smell, or an emotion from a feeling or the degree of touch, or the functions of the cerebral cortex from the cerebellum.


When a set of electrical impulses strikes at a set of chemical impulses, the sets of electrical impulses have motional dimensions that pull or push for these drifts, to appropriate knowing — and experiential outcomes.


The basic interaction processes of impulses are similar, the difference with the stairs or drift defines the complex consciousness of humans, with broad divisions and sub-divisions that include intelligence, memory, creativity, language, feeling, emotion, and so forth.


It is from the stairs that the sense of self also exists, for every process. Access to some of those drifts are the points of intentionality, free will, or control.


During controlled processes, prioritization happens briefly, before switching to pre-prioritization and back. Prioritized interactions define attention [or answering the binding problem].


It is postulated to be a result of the most rotation of a set of chemical impulses across the mind, at any instance.


There is a paper in Nature, Rotational dynamics reduce interference between sensory and memory representations, where the authors in the abstract stated that:

“Cognition depends on integrating sensory percepts with the memory of recent stimuli. However, the distributed nature of neural coding can lead to interference between sensory and memory representations.


Here, we show that the brain mitigates such interference by rotating sensory representations into orthogonal memory representations over time.


To study how sensory inputs and memories are represented, we recorded neurons from the auditory cortex of mice as they implicitly learned sequences of sounds.


We found that the neural population represented sensory inputs and the memory of recent stimuli in two orthogonal dimensions. The transformation of sensory information into a memory was facilitated by a combination of ‘stable’ neurons, which maintained their selectivity over time, and ‘switching’ neurons, which inverted their selectivity over time.”

Rotations Explained

This says that rotations separate sensory information and stored memories. Drifts or stairs, with their push or pull for sets of impulses — as earlier described — may sometimes appear like rotations fitting with this, but when sets of electrical impulses strike chemical impulses, there is a bit of rotation as well to determine what is prioritized.


Pain [and hurt] can be theoretically explained with, say, an ‘anticlockwise’ drift or rotation, whose rationing provides the experience for the sharpness felt.


What all impulses do when they interact is to drive knowing. There are features of impulses in sets that include pre-/prioritization, early-splits, sequences, principal spots, bounce points, and others.


Early-split explains how it is possible to see brain activities similar to a wakeful state during sleep, with an input, but to not have a response.


It means the incoming inputs were split, with some going-before but with a lower magnitude than regular, which could be because of some rotation but not comparable to when the individual is awake.


Organisms without a nervous system have a form of memory that does the functions of the mind for them. They can navigate and survive using a memory form that helps them to know.


AI does not have a nervous system, neither does it have cells like organisms, but it has machine memory that is dynamic enough to have outputs that compare to some of those by humans.


It may not have ‘experiences’ since it lacks sensory grounding, but some of what becomes of senses — or knowing — that humans use to relate with whatever comes in, is something it has.


LLMs already perform a vast aspect of knowing, making their functions vital to human consciousness. They may be seen as a tool to extend human consciousness, but humans may also present for them a completion of their lack of subjective experiences or other aspects of knowing.


Feature image source