Developing a Unique Breakthrough Algorithm in AI

In this research, we explored the development of a new AI algorithm that aims to address some of the current limitations in AI, specifically the lack of common sense and the absence of persistent mental representations. We proposed three potential solutions: a Common Sense Database, Learning Through Experience, and a Hybrid Approach. After a self-critique, we revised our approach to Learning from Text Data, which leverages the strengths of natural language processing (NLP) and machine learning to allow the AI to learn in a more flexible and adaptive way. This revised approach reduces the need for a large manually-encoded common sense database or a complex virtual environment. The research was conducted in a Jupyter notebook and involved the creation of pseudocode to outline the basic frameworks for each approach.