When Systems Thinking Means A Thinking System Then What ?
With the proliferation of seemingly thinking ( Ai ) systems how far can Ai system thinking + design go ? Far !
Past the discovery of of transformer models, CNN’s, Hybrid Neuro-Symbolic Systems ( LLMs + Symbolic Solvers ) used in reasoning in critical thinking applications, and onto Modular Multi-Modal + RAG designs, is where not only are specialized functional capabilities being stitched together in specific Ai models and to be able to seamlessly allow for intra-speciality spontaneity to emerge, it is also where evolutionary ( system design ) artifacts may emerge from exactly the same Ai systems. Is this recursive learning ? Does this mean systems that design themselves? Not in the way that is traditionally thought.
Where now, when Ai mavericks seemingly place ” simple ” processing ( as contrasted with say Ai drug discovery architectures to task ) what we see is, as an example in augmenting 3D design and modeling typologies with Ai, 3D tools like object multipliers, surface cloning, etc,. can even be cross purposed with standard innovation models of TRIZ and SCAMPER used for new product development efforts. It’s then where almost real time design, training and testing environments allow the mapping of unrelated functional tools. And, where this is no better described in Design Creativity in AI
When substitution, combination, adaptation, modification, repurposing, elimination, reversal; etc., efforts enhance ideation the amount of human validation necessary to move through the process of full say Ai system design, training and usage, dynamically, is slowly coming into an almost real-time environment. An Ai Vibe Architecture Interface.
This seemingly almost flies in the face of an evaluating criteria for ” value ” from investigations and where not only are there many such evaluating criteria for assessing creativity and alternative use cases, according to the López-Forniés system where Novelty, Usefulness, and Technical Feasibility may enumerate output, what are the ramifications of say using ( for example ) completely non visible spectrum analysis in say, auditory analysis is not something that can be quickly be created, tested and deployed in a rapid manner unless dynamic Ai system architectures can be used. And they are.