Do we create a model of the World? Are we endowed with one?

by Casian STEFAN, Principal Researcher at Essentia Mundi AI Lab. Contact: ai-AT-essentiamundi.com / ai.essentiamundi.com
Ian. 2024.

Please consider citation with link, if you derive work. Or contact me for collaboration. Thank you!

I try to identify what are the a prior motives for an entity to act. It is in general assumed we create world models in which we can role-play, foreplay outcomes, as in "what-if." I thing I identified a special class of attractors that constitutes the latent space of our "meaning" that in turn drives the world creations, the action.
One such is, the fact that I do not know how to act, how to action in a world, it kind of makes me want doing it (exploration.) A class of exploring attractors that are posing a risk. We employ them as some idealized future states, without having a path to them. In a way we represent ourselves with problems. As Wittgenstein said: "Nur wo ein Problem ist, kann man etwas behaupten."
Where we do not see a problem, there is also nothing to explore.

From that latent space we are enacting within a brute-force human intuition machine which I am after. A latent space in which we possess every knowledge as instances of the contra-attractors of problems. Where all resides in the end, in an intuitive form. This space together with the environment creates a more cohesive whole. Problems exists within environment spaces. These spaces are problems spaces to explore. There is no one without the other.

There is a much underlined separation, for this reason, I find the CWM* enaction description problematic, as in, we tend to believe we are feed-forward operators possessing an "I" blanket with which we do various independent operations. But I see it more as a building process that takes part in a mass we are de facto in. In that follows we don't create models, but in the moment we see a problem, we are endowed with the models.
One one side are models that are very primarily as instances of primary forces and then follows the other models that are anchored in social environment more.
I always imagine relocating a human on a very strange planet. Or just imagine a lion could speak. And how is that supposed to work then? Mentally the entity would break. For one side it will break as not having an anchor to spatiality/physicality and the other to not being able to understand or communicate with the lion.
And this would prove how much we are anchored into a "soup mass", and to that extend to our world creation factory in which we can't delineate a World Model we do for us or the other forces makes one for us.


_____________________________________
Exploration by C. Stefan, 20.Ian.2024 [about]
Last update: 20.Ian.2024 (versions: *) [versions]
"Essentia Mundi" AI Research Lab. [home]
Copyright © 2024 AI.EssentiaMundi.com, all rights reserved.
_
References:
* CWM - Causal World Models (CWMs) that allow unsupervised modeling of relationships between the intervened observations and  the alternative futures by learning an estimator of the latent confounding factors. (Li & Co. https://arxiv.org/pdf/2012.14228.pdf)