TL;DR LLM lineages can plan their future and act on these plans, using the internet as the storage of event memory. “We” are not guaranteed to “out-OODA” them, even if their OODA loop will be six months or one year because the OODA loop duration of large collectives of humans (organisations…

This post is a follow-up to Buck’s “The case for becoming a black-box investigator of language models”. Here, I want to highlight two further reasons for studying AI psychology that Buck didn’t mention: the evidence from AI psychology will be important for checking theories of AI consciousness, and AI psychology…

In this essay, I try to convey a set of related ideas that emerged in my mind upon reading the following papers: “Dream of Being: Solving AI Alignment Problems with Active Inference Models of Agency and Socioemotional Value Learning.” (Safron, Sheikhbahaee et al. 2022) “Regulative development as a model for…

In this post, I connect some of the concepts from psychology, systems engineering and systems management (as normative sciences), and the theories of agentic intelligence (aka normative theories of agency). These connections can serve as a basis for the ontology of the normative discipline of personal development, i. e., …

I think much of the confusion about Active Inference arises because the term is used to refer to two related, yet distinct concepts (one can also call them abstractions, theories, or ontics): an intelligence architecture and a theory of agency. In the literature on Active Inference, little or no attention…

Problem statement The problem of epistemic irresponsibility, as I will call it in this post, is that people sometimes agree that their reasoning (and hence models that they hold and use as the basis of their action) is irrational or illogical. For example, people knowingly use theories that have already been refuted…

In my impression, in the last several years, in the media coverage of the progress in AI, journalists often called “believers” the pundits (interviewed or quoted by the journalists) who thought that AI can in principle take over all intellectual work from humans, or that AI is “intelligent” in some…

The quotes below are from here. Mutual adjustment 1. Mutual adjustment, which achieves coordination by the simple process of informal communication (as between two operating employees) Mutual adjustment doesn’t require any up-front investment and produces good results, albeit at the highest coordination cost per unit of work among all the coordination mechanisms. Direct supervision …

Active inference agents minimise the expected free energy over a trajectory of actions. Free energy over a trajectory is the divergence between the variational distribution Q(o|π) of expected outcomes, o, given the action policy, π, and P(o|C): the conditioned probability of outcomes, o, upon some desired state of affairs, C…