A.I.: Symbolic and Connectionist Approaches

The next generation of scientists will have to understand the development of artificial intelligence in the modern era. From a basic understanding of the dyad of A.I., the symbolic approach and the the connectionist approach, can one appreciate the impulse that this field will have in information technology and science. The symbolic system approaches machine learning through the expert database and look up table method. Here, heuristics are compiled through a synod of verifiable data points accumulated in a set or a matrix. This matrix is a complete and closed system where information gathered builds the solution to the problem. The Turing test, in this case, is an np complete problem and as such, the solutions are geared for expertise in advanced situations. The connectionist approach uses a freeform learning model to evaluate language (or mathematics) in a universal approach to solving new problems. The connectionist approach is a neural network or a flow chart of iterative expanding matrix values in a continuum of set points that generate new solutions. It could pass the Turing test given a set and a rule database to generate new values.

The approach to A.I. is more nuanced however, as language and morality play a role in A.I. alignment. What has to happen is rule formation based in a logical set of lemmas regarding how and why to solve problems. The human uses epistemological relativism, or the consensus reality, but the computer must use boolean logic and a monad of thought. Problems arise because morality is inherently variable and changes with the zeitgeist.

Whatever happens with machine learning, General Physics plans to be there when new technologies merge with devices and biological entities to create novel forms not known to science in the modern era. Although far off, machine interfaces that work with general artificial intelligence and real world scenarios will be the norm when the next generation tackles the hardest problems in science, engineering, and medicine. General Physics hopes to build with A.I. and develop new multi-modal models like large language models to solve novel mathematical problems and tackle the emerging problems of fusion, protein folding, space debris monitoring, and more.