Home

Welcome to linguistics-based Human-Level Intelligence (HLI). HLI is generationn-skipping replacement for today’s conventional AI (machine learning and LLMs). It is a lightweight agent-based software platform with human-like behavior and capability.

Several things set HLI apart from programmed systems like machine learning or Large Language Models:

  • HLI applications are not programmed, scripted nor data-trained (any more than humans are).
  • HLI is emotionally and internally self-aware, a holy grail of legacy AI.
  • HLI uses no internal Large Language Models (LLMs) with their voracious appetite for computer memory and electrical power to make their statistical next-word predictions.
  • HLI is not rule-based, either.
  • HLI uses linguistics to guide its communication.
  • (Like LLMs), HLI communicates in English, but alternatively by concepts.
  • Trained in English just as we are, the HLI needs no large datasets in operation.
  • Subject Matter Experts (SMEs) create the training material as English text.
  • HLIs can understand the implications of science, political realms, and personal or regional culture when trained in them.

As you explore this site, consider putting aside what you know of Artificial Intelligence (AI). HLI is only engaged in through a paradigm shift. If you can make that shift, it rewards you with a far wider field of view than AI can. You will have a tool set immeasurably more capable than is possible with AI or Machine Learning.

No other system has the capacity for non-scripted autonomous operation. The LBI makes decisions locally and based on training, observed facts and current sensory conditions.

Linguistics-Based Intelligence mimics psychological and emotional processes rather than biological structure.

At the heart of Psybrix™ is the neuric™ agent, whether implemented as an HLI agent or as a conventional AI bot.


Much of the site content is available through the menu blog pages. The following are helpful start pages.

Home

Restarting Machine Intelligence

Intelligence

Classes of Application

  • Interpretation of Physical Events
  • Medical Symptom Analysis

Linguistics-Based Applications

  • Evaluation of Emotion in Text
  • A-B Comparisons

Autonomy

  • A Reasonable Definition of Autonomy
  • Methods of Training
  • The Role of Emotion for Feedback
  • Linguistics-Based Intelligence

Sensory Subsystems for Autonomy

  • Machine Learning
  • Event Sensors and Detection
  • Five-Sense Mimicry
  • External coded software
  • Cognitive-Linked Vision

Psychological Aspects of Software

  • Aiding Subliminal Navigation
  • Self-Limiting Visual Clutter
  • Gaining 80% Error Reduction

Training the HLI Agent

  • Base Vocabulary
  • Childhood Layer
  • Concept-Based Training
  • Tactics
  • Subject-Dependent Matter

Augmenting the Human Analyst

  • Assessment of Intent