Knowledge

The main value of a biological sensor is allowing the neural system behind it segregate separable ‘things’ in the world. In early stages, a kid normally learns that ‘things’ that are discovered through sensors are somehow reside outside. They belong to the world as opposed to being a part of ‘him/her’. We can roughly denote those ‘things’ as ‘physical objects’. For the majority of physical objects around us, vision is the main source of ‘truth’ about them. It seems that our other senses role is just supporting the knowledge we get through vision. Vision is highly correlated to ‘space’. ‘Space’ has several precise definitions in science; we use the common layman ‘meaning’ here. Audition on the other side correlates to ‘time’. Similarly; ‘time’ is something that science is heavily dealing with; we meant the ’time’ that everyone has the same intuition about. We know that vision is not happening instantaneously. It requires some time (fractions of a second) to stimulate the ‘retina’ or any other biological photo-sensor. However; this time is the shortest time among all sensors considering the total amount of information processed. The rate of processing visual stimuli is the fastest.

Can we draw a sharp line between the minimum acquirable ‘knowledge’ about physical objects and any concepts above? This is not an easy question. However, if we want to get extreme in minimizing that knowledge, we can define the ‘grounded’ knowledge that is acquirable from the sensors in the following way:

• Novelty: any novel sensation is regarded as new knowledge.
• Recognition: the ability to separate individual ‘things’ in a sensation sample (or scene).

Taking this tightly, if tow identical things are sensed in the same scene (space) or in different times have different color (or shade, orientation, size, ...) they will be recognized as different things. Recognition is basically based on contiguity (space/vision) or continuity (time/audition) of the stimulus.

Creatures like us can acquire knowledge of the world dynamics, using their neural systems (brains). While still directly connected to the ‘grounded knowledge’, a-little-higher concepts are incorporated. Following the same process, abstracted knowledge can be ‘developed’. The most amazing aspect of this development journey is that we invented body-extensions in the form of measurement apparatuses as extension to our bare sensory organs. At the same time, we develop tools and other engineering products as extensions to other parts of our bodies; considering their functions. This development process seems to go far beyond self-extensions which is true in some sense. An adjustment is proposed at the end of this article. The point I want to emphasize here is the ambiguous line between well-grounded knowledge and ‘conceptual’ knowledge.

Sensory organs are facing the outside world. Inner parts of our biological bodies are fairly complex ‘environments’ that share the same neural network. Inner stimulations and control actions start earlier than those connected to sensors. The also continue changing and evolving through our life time. They are a parallel source of knowledge although its mostly sub-symbolic. The most important point here is the natural knowledge mixing (or blending) process that take place at our brains. While marinating some degree of separation, the two sets of ‘knowledge’ are not entirely independent.

This ‘blending’ process provide a room for further abstraction. The brain dynamics allow the development of totally new breed of concepts and ideas on top of the ‘mix’. Again we can’t draw a sharp line between the knowledge that is built on just the grounded knowledge and that make use of the internal body dynamics as well. However, we can easily attribute things like ‘creativity’ and ‘intuition’ (partially) to this inside-outside blending process.

The idea of looking at the biological body as a ‘universal’ antenna add a fifth kind of knowledge to the four illustrated ones. The mechanism is not some magical knowledge receiver that is somehow tuned to all natural sources of ready-made knowledge. Its rather the degree of exposure that allow every single cell of our bodies to have experiences beyond the internal and external stimulations that we already know.

As a summary; knowledge can casually be categorized into five distinct categories:

1. Grounded knowledge
2. Conceptual knowledge
3. Inner knowledge
4. Creative knowledge
5. Spiritual knowledge

I think the current AI stage is mostly at (1). Even with LLMs; if you consider their ‘world’ to be ‘words’, they do full job on (1) and a little leap to (2). AI is something different than biological us. That means to fulfil its objectives (the companies’ objectives!!), they are free to focus on a subset of knowledge categories or even to introduces new ones.