Peter Voss - Essentials of General Intelligence: The Direct Path to Artificial General Intelligence (2007)

History / Edit / PDF / EPUB / BIB /
Created: March 3, 2016 / Updated: November 2, 2024 / Status: finished / 7 min read (~1375 words)
Artificial General Intelligence

  • Intelligence can be defined simply as an entity's ability to achieve goals - with greater intelligence coping with more complex and novel situations
  • General intelligence comprises the essential, domain-independent skills necessary for acquiring a wide range of domain-specific knowledge
  • Learning must be autonomous, goal-directed and adaptive
  • The mark of a generally intelligent system is not having a lot of knowledge and skills, but being able to acquire and improve them and to be able to appropriately apply them
  • An AGI system should be able to learn to recognize and categorize a wide range of novel perceptual patterns
  • It should be able to autonomously learn appropriate, goal-directed responses to input contexts (given some feedback mechanism)

  • Perceived entities/patterns must be stored in a way that facilitates concept formation and generalization
  • An effective way to represent complex feature relationships is through vector encoding
  • Any practical applications of AGI must inherently be able to process temporal data as patterns in time - not just as static patterns with a time dimension
  • AGIs must cope with data from different sense probes and deal with attributes such as: noisy, scalar, unreliable, incomplete, multi-dimensional, etc.
  • Another essential requirement of general intelligence is to cope with an overabundance of data
  • The system needs to have some control over what input data is selected for analysis and learning - both in terms of which data, and also the degree of detail
  • Senses are needed not only for selection and focus, but in order to ground concepts in reality

  • In addition to understanding general intelligence, AGI design also requires an appreciation of the differences between artificial (synthetic) and biological intelligence, and between designed and evolved systems
  • Work focused on
    • General rather than domain-specific cognitive ability
    • Acquired knowledge and skills, vesus loaded databases and coded skills
    • Bi-directional, real-time interaction, versus batch processing
    • Adaptive attention (focus and selection), vesus human pre-selected data
    • Core support for dynamic patterns, versus static data
    • Unsupervised and self-supervised, versus supervised learning
    • Adaptive, self-organizing data structures, versus fixed neural nets or databases
    • Contextual, grounded concepts, versus hard-coded, symbolic concepts
    • Explicitly engineering functionality, versus evolving it
    • Conceptual design, versus reverse-engineering
    • General proof-of-concept, versus specific real applications development
    • Animal level cognition, versus abstract thought, language, and formal logic
  • Self-improvement takes two distinct forms/phases:
    • Coding the basic skills that allow the system to acquire a large amount of specific knowledge
    • The system reaching sufficient intelligence and conceptual understanding of its own design, to enable it to deliberately improve its own design
  • Many AI systems do all of their learning in batch mode and have little or no ability to learn incrementally
  • In many cases they are unable to adapt beyond the initial training set without reprogramming or retraining
  • Intelligent systems must be able to act
    • Acting on the "world" - be it to communicate, to navigate or explore, or to manipulate some external function or device in order to achieve goals
    • Controlling or modifying the system's internal parameters (such as learning rate or noise tolerance, etc.) in order to set or improve functionality
    • Controlling the system's sense input parameters such as focus, selection, resolution (granularity) as well as adjusting feature extraction parameters
  • AGI systems must inherently be designed to acquire knowledge by themselves
  • They need to control what input data is processed, where specifically to obtain data, in how much detail, and in what format
  • All acquired knowledge and skills is encoded in one integrated network-like structure
  • One can say that "high-level intelligence is conceptual intelligence"
  • Autonomous concept formation is one of the key tests of intelligence
  • Design to achieve the desired functionality of the brain rather than try to replicate evolution's design
  • Here is a list of desirable cognitive features that can be included in an AGI design that would not exist in a reverse-engineered brain:
    • More effective control of neurochemistry (emotional states)
    • Selecting the appropriate degree of logical thinking versus intuition
    • More effective control over focus and attention
    • Being able to learn instantly, on demand
    • Direct and rapid interfacing with databases, the Internet and other machines - potentially having instant access to all available knowledge
    • Optional "photographic" memory and recall on all senses
    • Better control over remembering and forgetting (freezing important knowledge, and being able to unlearn)
    • The ability to accurately backtrack and review thought and decision processes (retrace and explore logic pathways)
    • Patterns, nodes and links can easily be tagged (labeled) and categorized
    • The ability to optimize the design for the available hardware instead of being forced to conform to the brain's requirements
    • The ability to utilize the best existing algorithms and software techniques - irrespective of whether they are biologically plausible
    • Custom designed AGI can have a simple speed/capacity upgrade path
    • The possibility of comprehensive integration with other AI systems (like expert systems, robotics, specialized sense pre-processors, and problem solvers)
    • The ability to construct AGIs that are highly optimized for specific domains
    • Node, link, and internal parameter data is available as "input data" (full introspection)
    • Design specifications are available (to the designer and to the AGI itself!)
    • Seed AI design: A machine can inherently be designed to more easily understand and improve its own functioning - thus bootstrapping intelligence to ever higher levels
  • Discoveries in cognitive psychology point towards generalized pattern processing being the foundational mechanism for all higher level functioning

  • General intelligence requires a number of foundational cognitive abilities:
    • Remember and recognize patterns representing coherent features of reality
    • Relate such patterns by various similarities, differences, and associations
    • Learn and perform a variety of actions
    • Evaluate and encode feedback from a goal system
    • Autonomously adjust its system control parameters
  • Pattern acquisition through lazy learning
    • Stored feature patterns with adaptive fuzzy tolerances
    • Recognition/Pattern matching through a competitive winner-take-all, as a set or aggregate of similar patterns, or by forced choice
  • The matching algorithm is able to recall patterns by any dimension

  1. Development framework
  2. Memory core and interface structure
  3. Individual foundational cognitive components
  4. Integrated low-level cognition
  5. Increased level of functionality
  • AGI engine with the following basic components:
    • A set of pluggable, programmable (virtual) sensors and actuators (called probes)
    • A central pattern store/engine including all data and cognitive algorithms
    • A configurable, dynamic 2D virtual world, plus various training and diagnostic tools
  • Additional details:
    • Data recorder with playback
    • Data visualization and editing tools
    • A cognitive core with many foundational cognitive algorithms
    • An interface manager which communicates with the probes, the cognitive core and the data recorder

  • Can be separated into three parts:
    • Cognitive core
    • Control/Interface logic
    • Input/Output probes
  • Cognitive core
    • Central repository of all static and dynamic data patterns - including all learned cognitive and behavioral states, associations, and sequences
    • All data is stored in a single, integrated node-link structure
  • Control and interface logic
    • Coordinates the network's execution cycle, drives various cognitive and housekeeping algorithms, and controls/adapts system parameters
    • Via an interface manager, communicates data and control information to and from the probes
  • Probes
    • Programmable feature extractors, variable data resolution, focus and selection mechanisms
  • Development environment, language, and hardware
  • Implemented in C#/.NET
  • Practical/Proof-of-concept prototype performance can be achieved on a single PC (2 GHz, 512 MiB)

  • General intelligence emerges from the synergetic integration of a number of essential fundamental components

  • Classifies their work in the area of agent systems and embodied cognitive science

  • Some technical issues worth mentioning:
    • Epistemology: Theory of knowledge, the nature of knowledge, and how it relates to reality.
    • Theory of mind: The formulation and understanding of consciousness, intelligence, volition, meaning, emotions, common sense, qualia.
    • Cognitive psychology: Proper understanding of the concept intelligence.
    • Project focus: A vision of how to get from here to there.
    • Research support
      • Incremental, real-time, unsupervsed/self-supervised learning
      • Integrated support for temporal patterns
      • Dynamically-adaptive neural network topologies
      • Self-tuning of system parameters, integrating bottom-up (data driven) and top-down (goal/meta-cognition driven) auto-adaptation
      • Sense probes with auto-adaptive feature extractors
    • Cost and difficulty: Finding the crucial fundamental functionality. Scaling up the system to human-level storage and processing capacity