ActInf GuestStream 084.1 ~ "The Nature of Habits and Agential Systems", Jesse G

[@jesseg7841]

0:00:00 - 0:13:14 Introducing content and self-introduction
0:13:14 - 0:17:44 Stepping through the paper... Understanding habits and the nature of agency; reinforcement learning as opposed to maladaptive learning
0:17:44 - 0:18:36 Perceptions... Actions... Priors... denoting the PAP-cycle and the critically existential nature of stimulus...
0:18:36 - 0:21:43 Single Agent Systems vs Multi-Agent Systems... focusing on the contextualization of Multi-Agent Systems
0:21:43 - 0:23:15 Classical Computer Systems vs Agential Systems... and the fundamental nature of stimulus being critical for agential systems...
0:23:15 - 0:28:36 Elaborating on Perception in PAP-cycle along with generalization of sensations to be filtered into being prioritized as generalization of information
0:28:36 - 0:31:23 The problem of the capacity of agents to resolve issues with perception... understanding veridical perception limits
0:31:23 - 0:32:57 Elaborating more on the nature of habits and their critical role in regulating agency for agential systems... nature vs nurture dynamics... intrinsic vs extrinsic variability
0:32:57 - 0:35:12 metabolic load vs cognitive load... thinking bottom-up as opposed to top-down dynamics regulating the integrity of agents
0:35:12 - 0:37:57 Elaborating more on nature vs nurture dynamics
0:37:57 - 0:39:20 Dynamics of metabolic load being fundamental to all possible behaviors enacted by agents
0:39:20 - 0:41:43 Routines being fundamental to the regulation of habits
0:41:43 - 0:43:47 Regarding the nature of environment and agents being dynamical and never in a fixed static state... regarding the nature of complex adaptive systems
0:43:47 - 0:44:13 Intrinsic vs extrinsic variability in terms of regularity
0:44:13 - 0:46:30 More elaboration on reinforcement learning as opposed to maladaptive learning... regarding tug-of-war transient dynamics
0:46:30 - 0:51:09 Dynamics of goal-directed oriented behavior being connected to habits
0:51:09 - 0:53:36 Dynamics in which agents wrestle to regulate their optimal integrity... or fitness... in the face of entropy
0:53:48 - 0:56:31 Resources as opposed to obstacles within the embedded environment in which agents are existentially bound to
0:56:31 - 1:01:08 Dynamics of homeostatic regulation being central to fitness with respect to guided random search optimization behavior... introducing the idea of axiomatic nature of behavior
1:01:08 - 1:08:33 Elaborating on Actions and Priors as opposed to Perceptions with regard to the PAP-cycle... and more on prioritizing generalizations of sensations into information
1:08:33 - 1:13:10 Back to guided random search optimization... elaborating more on guided and search aspects as opposed to randomness
1:13:10 - 1:16:58 Homeostatic Regulation being critical to agents regulating their integrity
1:17:16 - 1:19:09 Summarizing homeostatic regulation with respect to guided random search... summarizing the nature of stimulus being anything that provokes an agent to engage intrinsically or extrinsically... summarizing how the environment presents resources and obstacles for agents to regulate themselves
1:19:09 - 1:30:05 Mechanistic interpretability explanation of the PAP-cycle... regarding sensor and actuator dynamics of neural and glial cellular networks... tying it back to metabolic load (bottom-up) as opposed to cognitive load (top-down)...micro level as opposed to macro and meso levels of regulation...
1:30:05 - 1:35:24 Hierarchical vs Heterarchical processing structure of the central nervous systems of humans as multi-agent systems... prioritization as opposed to selective tuning parameter mechanisms
1:35:24 - 1:38:12 introspective and interoceptive as opposed to proprioceptive capacities regulating the PAP-cycle
1:38:18 - 1:41:56 Small summary regarding regions of the brain responsible for coordinating PAP-cycle dynamics
1:42:18 - 1:45:54 More elaboration on generalizations of sensations... elaborating more on metabolic load being a visceral process that is constantly happening as the agent continues to exist
1:45:54 - 1:54:12 System 1 and System 2 coupled with The Rider and The Elephant... reasoning interwoven with emotions (affective processing)... understanding embodied cognition
1:54:12 - 1:55:16 Re-summarizing hierarchical as opposed to heterarchical processing
1:55:16 - 2:07:37 Emphasizing that reality is an interface... more regarding existential embedding of agents within the environment they are bound to exist in
2:07:37 - 2:14:11 Preserving viability... regarding inter-agent cooperation dynamics
2:14:11 - 2:33:39 Dynamics in which agents handle moments of uncertainty... interpretive vs functional-based indeterminism
2:34:10 - 2:38:17 Elaborating on reinforcement learning regarding the role of deterrence
2:38:17 - 2:42:00 Regarding reinforcement learning with regard to inter-agent cooperation due to sophisticated language structure utility (QUICK SUMMARY)
2:42:00 - 2:44:55 Elaborating on maladaptive learning regarding cognitive strain (QUICK SUMMARY)
2:44:55 - 2:46:30 Free Will dynamics... regarding the degrees of freedom agents have with respect to their engagement and overall fitness (QUICK SUMMARY)
2:46:30 - 2:54:21 Concluding remarks

[@ryanjbuchanan]

This was a very interesting watch. Jesse is a cool guy and I really enjoy his research