Discourse does not allow reuse of topic titles…
Whilst I don’t much care for the term “cognitive” in the context of computing, I found a piece linked with a phrase the same as my subject line:
and that link came from their github repo:
which appears to be a community version of:
which is a link I got from @BentleyDavis in his post at the CDL slack.
That piece builds on an earlier piece
dated due to truly stale hyperlinks, but the concepts serve, IMHO, as a rich introduction to the spaces of knowledge representation (KR) and information-sharing protocols, a topic related to @peterkaminski’s post on knowledge graphs.
Their objectives are important to FJB and OGM:
Describe a general framework that could be used to abstract cognitive processes into a network-graph model.
Based on this framework, design a scalable data model that could be used to store, retrieve, and discover new knowledge: in short, which could emulate all the facets of thinking process.
That last sentence fragment is where we get allusions to the idea that a computer can mimic human cognition. My view: at best, computers can augment human cognition.
The conceptual model, given this is a 2014 post, represents nothing new, but remains a concise overview of the overall space in which we use computing to augment human capabilities:
- The basic building block of the model is a concept
- The concepts appear within statements
- The statements appear within contexts
- The concepts , the statements , and the contexts are all made by users
- The narrative provides a way of representing the statements and concepts in a sequential order
- The user is either the one who makes the statements or the one who receives them.
The last point brings to mind Gordon Pask’s Conversation Theory, in which speakers hold domain models as well as models of their listener’s capabilities, and listeners hold similar models; the action going on in a conversation exists in a cloud between the two, which Pask called an entailment mesh - today, we might say knowledge graph; Conversation Theory posits that conversations are about negotiating the structure of that graph.
So, maybe, if we blend Pask’s ideas, particularly the notion that our conversations are about negotiating a graph which aims at mutual understandings, then maybe we can allow the term “cognitive” to creep in, if only for conversational purposes. As Wikipedia says:
Conversation theory describes interaction between two or more cognitive systems, such as a teacher and a student or distinct perspectives within one individual, and how they engage in a dialog over a given concept and identify differences in how they understand it.
Narrative ties all of this back to the recent explosion of banter in OGM about storytelling. That is, in the universe I am trying to sketch here, everything starts with storytelling, but goes way beyond that, beginning with the ways in which we use those stories to derive deeper and mutual understandings. Graph models appear to be crucial to the ways in which we will use computing in that process.