In Search of Artificial Intelligence with Human Intelligence

Reasoning, problem solving, and learning are the crucial facets of human intelligence. Without understanding human intelligence or the logic behind human communication, there is little hope that the current science will advance humanity much more than what is possible. To break this barrier, we need to reconcile the relationship between human innate abilities such as creative thinking, critical thinking, language, perception, semantics and the science behind computational epistemology.

Automation has contributed to much of the advances in manufacturing, quality and economic value, but it is also restraining human thinking to repeating procedures.

What's the economic value of understanding narrative at scale without misreading?

Natural language is created for human communication. Human intelligence is good at understanding the context and semantics of written narratives, but falls short in doing it at scale to avoid fallacies, conjecture, and misleading statements. Worse yet, when Generative AI is used to summarize a narrative for important decisions. Generative AI does not understand context based on semantics, instead, it uses token pattern derived from text and calls it context. It is a source of misinformation when Generative AI generated summary is being treated as a correct summary of natural language documents.

Economically, misunderstanding of context and semantics in critical documents such as legal works, contracts, statements of work, sales transcripts, meeting minutes, opinions, earning transcripts, or business intelligence will result in unwanted outcome from poor decisions.

What's missing in current Science?

Existing LLM transformer introduces uncertainty and nonsense without logic or semantics, making it useless in most cases.

What is missing in the current science is the connection between semantics, logic and mathematics driven by human consciousness and perception.

Abstractive Symbolic AI is a technology that uses informal logic together with computational epistemology and symbolic logic to analyze written narratives, resulting in deterministic interpretation of semantic relevance, headwind, tailwind and author's perspective.

That's our implementation of ELAINE - Abstractive Symbolic AI.

ELAINE - Abstractive Symbolic AI

SiteFocus has developed a system that is not based on algorithm, statistics, training data or hard coded rules. The concept of "semantic hierarchy" as a framework for narrative is not hard to understand. Understanding this principle will help professional to resolve many difficult problems.

As a start, you can explore the power of "semantic hierarchy" with examples of ELAINE analysis in the ELAINE JOURNAL section. Articles of current events and research papers are analyzed daily. If you are interested to learn more about this technology, send us a request via email to explore how ELAINE can be applied to your business.