2018: the year Digital Transformation meets Artificial Strategic Intelligence

What is missing in today’s AI?

Today’s Artificial Intelligence (AI) based on big data, machine learning, and deep learning is great approach to discover things from the past and to improve precisions. This includes advantages driven mainly by integrating automation into products and services that derived from trends, consumer behavior, translation, association of patterns, resolving images and many other applications in consumer products. What’s missing in this big picture is intelligent and predictive capabilities. As much as a large portion of big data analytics is about predictive analytics, these analytics is predicting based on regression, such as predicting outcome base on past events. However, current events such as Bitcoin and 2017 Tax Reform are not prescribed by historical data, and therefore is beyond the scope of these analytics. Yet, there are other important forces that drive our life exerted by the force of nature or man-made forces. We can classify the Tax Reform as man-made forces, and hurricane as force of nature. These forces deliver huge impact to our life and economy, and ultimately affect the bottoms line on business earnings in a big way.

A new technology driven by need

There is another branch of AI that is less visible these days which helps to predict outcomes of things and events such as Bitcoin, 2017 Tax Reform and media engineering. It is Artificial Strategic Intelligence — a term that popular media seldom talk about and is not generally found in consumer goods and services. You may have heard about it being discussed as rationale based on the “force of nature” and “force of human behavior”. It is the encapsulation of symbolic logic based on behavior of people, climate, environment, economy, culture, things of nature and man-made. One example of behavior of people is sentiment analytics. This kind of analytics is common among Natural Language Processing providers. We call this “Artificial Strategic Intelligence”. The reason is because this type of AI is mainly used to help decision makers to form strategies for businesses and governments. Most C-Suite officers will benefit from this type of AI. As much operational metrics from quantitative analysis is an important part of decisions, quantitative analysis alone cannot provide insights into things and events like Bitcoin, 2017 Tax Reform, or global supply chain issues.

Selecting technology for implementation

Artificial Strategic Intelligence sounds profound but in practice it is not difficult to acquire or manage if it is done right. The challenge in developing this type of AI is the effort required to build a new technology from the ground up that is capable of understanding the unknown. Prevailing technologies such as machine learning or deep learning cannot be used since these technologies are made for known patterns. Meeting this challenge require a full stack of skill to refine the symbolic logic and the implementation of a system that cannot take advantage of much of the open-source frameworks. For examples:

  • Many existing frameworks such as Hadoop, MapReduce, Spark, Storm, Kafka are great open-source for big data processing, but not much help in symbolic processing.
  • Lucene, Elastic Search, Solr, and Rdis are great for symbols and cache, but lack in symbolic logic manipulation.
  • Machine learning libraries such as Mahout, Spark ML and TensorFlow are great for statistical learning, but not much help in symbols and legible rules learning.
  • Relational databases and NOSQL are great for structures based on entities and relationships, but not much help in discovering symbol entities and symbol relationships.
  • Other big data tools such as Cassandra, MongoDB, or CouchDB are great tools for huge repositories, not much help in the manipulation of relationships between unstructured data.

However, there are many tools and data structures in JAVA that forms the fundamental building blocks of the above tools and are supported by many design paradigms such as micro-service, multi-threading, M2M, HPC, MPP, distributive, container and orchestration. As a modern computer language, JAVA is both a functional language in Lambda, and object oriented programming with powerful interfaces for multi-inheritances and new technology adoption. As a bonus, JAVA can seamlessly integrate with other languages such as C or C++. JAVA becomes our language of choice for this new AI paradigm.

Communication in Focus (CIF) — putting Artificial Strategic Intelligence to work

Our effort to build an AI system from the ground-up was not a simple task. Picking the right tool was a good start. For building blocks, we identified the following components that formed the core of our Artificial Strategic Intelligence product — CIF:

  • Point of View Analytics that emulate focus groups and depicts human behavior
  • Bionic Fusion enables man-machine integration
  • Context Discriminant explores natural language by contexts and subjects
  • Meta Vision and Knowledge Vision enhance the effect of Bionic Fusion
  • Transient Context Discriminant enables refinement of Meta Vision on subjects
  • Context Discriminant Calculus derives Semantic Neighborhoods to a higher order
  • Web Interfaces with Micro Services implement Lambda Architecture
  • Distributive Network manages multiple data-lakes and asynchronous jobs
  • Context Aware Web Bot acquires content from the Internet
  • Visual display of Meta Vision and Knowledge Vision renders visual analytics
  • A collaborative environment enables sharing and off-session collaboration
  • CIF Content Ingestion provides instant data ingestion on text, PDF, DOC, DOCX, and URL
  • A simple, yet profound repository connects entity, project, and topic relationships
  • Multi-Aspect NLU Report Generator generates on-demand interactive insight reports
  • A role based access control Web Interface ensures security and multi-tenancy
  • Web Service enables Machine to Machine interaction

Finally, our system was put together for a benchmark with the following topics:

  • Enron Email Corpus was analyzed and produced a Multi-Aspect NLU Report in less than 50 minutes. This report depicts every aspect of events leading up the debacle of Enron and subsequent legal entanglements. It demonstrates how this technology can be used in conjunction with SOX (Sarbanes-Oxley) in compliance management, internal control, governance, and government regulations.
  • Daily concurrent news analytics published with Multi-Aspect NLU Report reflecting the AM, PM, and special financial news events
  • Multi-Aspect NLU Report on SEC filings on IPO Prospectus (S1) and M&A filings (S4)
  • Enterprise earning call analysis using Multi-Aspect NLU Reports and Meta Vision reconciles executive statements with analyst Q&A
  • Analytic reports detect effect of media engineering by enterprise
  • Competitive BI reports based on enterprise products and service

Our 2018 business objective is charted to provide business leaders and C-Suite officers with this new AI tool for in-time deployment with their strategic planning.