On Sunday, September 05, 2021, at 5:11 AM, Nanograte Knowledge Technologies wrote: > For historical references, Common Lisp and even Lisp 1.5 had reflection too. > Around 1980.
> (On my spare time, with others, I am slowly developing http://refpersys.org/ > - an open source GPLv3+ licensed *ref*lexive *per*sistent *sys*tem for Linux > - which could interest some people on this list) On my path in computer programming I haven't done Lisp but plan on looking at Clojure… C# has the following features as do other languages: reflection, anonymous types/dynamic types, delegates (function pointers), serialization, and with these many things can be done. Though the syntactic expressivity of reflective semantics in C# leaves much to be desired. That’s probably why its reflection hasn’t been fully taken advantage of. But it works… I’ve thought about how to have proto-AGI code write itself in C#. What I’m thinking is something like: Reflecting reality by environmental dynamical systems modelling, continuous model checking contributing to self-improvement, learned systems absorption into mathematical structure, reflecting the application itself with consciousness centered on app execution context, a dynamical, self-adapting, self-writing test system, then app self-organizingly forms and unfolds into its existing precepted reality reflecting itself into its own environmental systems modelling. A question is should a layer above C# reflection be added, a structure that is more syntactically aligned with reflective conceptual constructs, including the other language features mentioned above. Basically new object structure where object members have more syntactic expressiveness. OOP has its limitations… I think the key is in the delegate handling with the object property processing and network structure created off of that. For example, increasing distance pressure between two dynamic objects but keeping them linked by creating a growing graph of objects connected by dynamic member creation. A multigraph node is a C# object with connectors as member object references. Hypergraphs then are inherited objects. Node detection is reflection and subgraph isomorphism mining uses reflective processes. ------------------------------------------ Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/Tace3f9aea35af378-Me1427eccf6eebcdee5713d59 Delivery options: https://agi.topicbox.com/groups/agi/subscription
