Artificial Intelligence Systems Integration: Fundamentals and Applications

By Fouad Sabry

What Is Artificial Intelligence Systems Integration

Making discrete software components, such as voice synthesizers, compatible with other components, such as common sense knowledgebases, is the core concept behind the integration of artificial intelligence systems. The goal of this endeavor is to develop artificial intelligence systems that are larger, more comprehensive, and more powerful. The primary approaches that have been suggested for integration are known as message routing or communication protocols. These protocols are utilized by the software components in order for them to connect with one another, typically by means of a middleware blackboard system.

How You Will Benefit

(I) Insights, and validations about the following topics:

Chapter 1: Artificial Intelligence Systems Integration

Chapter 2: History of Artificial Intelligence

Chapter 3: List of Artificial Intelligence Projects

Chapter 4: Symbolic Artificial Intelligence

Chapter 5: Hybrid Intelligent System

Chapter 6: Cognitive Computing

Chapter 7: Glossary of Artificial Intelligence

Chapter 8: Situated Approach (Artificial Intelligence)

Chapter 9: Outline of Artificial Intelligence

Chapter 10: OpenCog

(II) Answering the public top questions about artificial intelligence systems integration.

(III) Real world examples for the usage of artificial intelligence systems integration in many fields.

(IV) 17 appendices to explain, briefly, 266 emerging technologies in each industry to have 360-degree full understanding of artificial intelligence systems integration' technologies.

Who This Book Is For

Professionals, undergraduate and graduate students, enthusiasts, hobbyists, and those who want to go beyond basic knowledge or information for any kind of artificial intelligence systems integration.

• Language: English
• Publisher: One Billion Knowledgeable
• Release date: Jun 23, 2023

Book preview

Chapter 1: Artificial intelligence systems integration

Making specific software components, like voice synthesizers, compatible with other components, like common sense knowledgebases, in order to create larger, broader, and more powerful A.I. systems, is the fundamental concept of artificial intelligence systems integration. The primary approaches of integration that have been suggested involve message routing, or the communication protocols that the software components use to speak with one another, frequently via a middleware blackboard system.

The majority of artificial intelligence systems use a combination of technologies, for example, the combination of speech recognition technology with speech synthesis technology.

However, in the recent past, The significance of systems integration as a distinct field has come under increased dispute.

Those who favor this strategy include academics like Marvin Minsky, Aaron Sloman, Deb Roy, Kristinn R.

Thórisson and Michael A.

Arbib.

a justification for the recent interest A.I.

Integration is appealing because some (relatively) straightforward A.I. systems have already been developed.

systems for particular issue areas (like computer vision), speech synthesis, etc.), and that a more sensible strategy for developing wider A.I. is integrating what is already available.

as opposed to creating monolithic systems from start.

The emphasis on system integration, particularly in relation to modular techniques, results from the fact that the majority of significant scale intelligences are made up of a variety of processes and/or make use of multi-modal input and output. A humanoid-type of intelligence, for instance, would ideally need to be able to speak using speech synthesis, hear using speech recognition, understand using a logical (or other unspecified) method, and so on. Integration of various modalities is required to create artificially intelligent software with broader intelligence.

The size of software firms and the size of their software departments is proof that collaboration is a crucial component of software development. Various protocols and guidelines that developers can adhere to ensure quality, dependability, and that their software is interoperable with software generated by others are among the tools to facilitate software collaboration (such as W3C standards for webpage development). Collaboration in the field of artificial intelligence, however, has generally been absent and confined to reputable academic institutions (and sometimes not within them either). This poses a significant challenge for those who work in artificial intelligence systems integration, and it frequently forces researchers to re-invent the wheel every time they want a particular functionality to operate with their program. The not invented here mentality, which shows up in a significant resistance on the part of A.I. researchers to build on the work of others, is much more destructive.

There are many solution islands as a result of this in artificial intelligence: Numerous discrete software elements and methods that handle distinct aspects of intelligence separately have been developed as a result of AI research. To provide a few instances:

Speech synthesis

FreeTTS from CMU

Speech recognition

Sphinx from CMU

Logical reasoning

OpenCyc from Cycorp

Open-Minded and Reasonable From MIT, net

Numerous pieces of newly developed software, including artificial intelligence (AI) systems, are open to public exploitation as a result of the free software movement's rising popularity. The logical next step is to combine these many software parts into cohesive, intelligent systems of a more general character. Since the community has already produced a large number of components (many of which serve the same goal), the easiest approach to integrate them is to give them all a simple means of communication. Each component becomes a module as a result, allowing it to be tested in different settings and configurations of bigger structures.

There are many online communities for AI developers where tutorials, examples, and forums are intended to aid both novices and seasoned engineers in creating intelligent systems. However, few communities have been able to successfully popularize a particular standard or code of behavior to make it possible for the vast array of unrelated systems to be merged with any ease.

A formal methodology called the Constructionist design methodology (CDM, sometimes known as Constructionist AI) was put forth in 2004 for use in the creation of communicative humanoids, cognitive robots, and large-scale AI systems. In order to produce coherent system behavior, such systems must include a multitude of features that must be carefully coordinated. The foundation of CDM is an iterative design process that produces a network of named interacting modules that communicate with one another via explicitly written streams and discrete messages. The CDM served as the model for the OpenAIR messaging protocol, which has been used frequently to support the creation of intelligent systems employing the CDM (see below).

ASIMO, a humanoid robot from Honda, and QRIO, a humanoid robot from Sony.

Cog, a humanoid robot project at M.I.T. led by Rodney Brooks.

AIBO, Sony's robotic dog, combines its senses of sight, sound, and movement.

Robot TOPIO, created by TOSY, can compete against people in ping-pong.

{End Chapter 1}

Chapter 2: Physical symbol system

Taking physical patterns (symbols), integrating them into structures (expressions), and then manipulating them (using processes) to generate new expressions is the function of a physical symbol system, which is also known as a formal system.

Allen Newell and Herbert A. Simon are the ones responsible for developing the physical symbol system hypothesis (PSSH), which is a philosophical perspective in the field of artificial intelligence. They penned it:

A physical symbol system has the essential in addition to the enough means for broad intelligent action.

— Allen Newell and Herbert A.

Simon

This assertion suggests not only that human thought is a kind of symbol manipulation (because intelligence requires the presence of a symbol system), but also that computers are capable of exhibiting intelligent behavior (because a symbol system is sufficient for intelligence).

The theory has been subjected to severe scrutiny from a number of different perspectives, despite the fact that it is an essential component of AI research. The notion that the theory appears to be acceptable for higher-level intelligence, such as playing chess, but less fit for ordinary intelligence, such as vision, is one of the more prominent critical perspectives. A line of demarcation is typically drawn between the kinds of high-level symbols that directly correspond with objects in the real world, such as dog and tail, and the more complex symbols that are present in a machine, such as a neural network. This distinction is made because high-level symbols like dog and tail directly correspond with real-world objects.

The following are some examples of physical sign systems::

Formal logic: the symbols are words like and, or, not, for all x and so on. These expressions are assertions of formal logic, and like other statements in logic, they may be either true or untrue. The procedures are the guidelines for inference and deduction in logic.

In algebra, the symbols are the plus sign, ×, x, y, 1, 2, 3, etc.

Equations may be derived from the expressions.

The operations are the fundamental laws of algebra, that enable one to modify a mathematical statement while preserving the expression's truth.

A digital computer, in which the processes are the actions of the central processing unit (CPU) that modify memory and the symbols are the ones and zeros that make up the computer's memory.

Chess is represented by three types of elements: the pieces, the processes, and the expressions. The pieces represent the symbols, while the processes represent the lawful chess movements.

According to the physical symbol system theory, both of these may be categorized as exemplifications of physical symbol systems:

Intelligent human mind is shown by the encoding of the symbols in our brains. The expressions represent different ways of thinking. The processes are the activities of the mind involved in thinking.

A program for artificial intelligence that is now being executed; the symbols represent data. The phrases constitute additional pieces of data. The processes are the programs that are responsible for the data manipulation.

The development of artificial intelligence programs and psychological research on human beings provided Allen Newell and Herbert A. Simon with two lines of evidence suggesting that symbol manipulation was at the core of both human and machine intelligence.

First, in the early decades of artificial intelligence research, there were a number of very successful programs that used high level symbol processing. Some examples of these programs include Newell and Herbert A. Simon's General Problem Solver and Terry Winograd's SHRDLU. Both of these programs were created by Newell and Herbert A. Simon. This heritage is carried on by its offspring, namely logic programming and expert systems. It was indicated by the effectiveness of these programs that symbol processing systems might replicate any intelligent activity.

And second, psychological investigations that were being conducted at the same time revealed that individuals employed this kind of symbol processing for complex issues in logic, planning, or any form of puzzle solving. This was discovered at the same time as the first point. Researchers working in the field of artificial intelligence were able to successfully model the process by which individuals solve problems using computer systems. This partnership, along with the questions it posed, would ultimately result in the birth of the academic discipline known as cognitive science. (The researchers gave this kind of investigation the name cognitive simulation.) Based on the findings of this body of research, it was hypothesized that human problem solving essentially consisted of the manipulation of high-level symbols.

In the arguments of Newell and Simon, the symbols that the hypothesis is referring to are physical objects that represent things in the world, symbols such as dog that have a recognizable meaning or connotation and can be composed with other symbols to create more complex symbols. In other words, symbols are physical objects that represent things in the world.

On the other hand, one may understand the hypothesis as referring to the basic abstract 0s and 1s that are stored in the memory of a digital computer, or to the stream of 0s and 1s that is processed by a robot's perceptual apparatus. Although it is not always easy to discern precisely what the symbols stand for, they may also be thought of as symbols in some sense. However, it may not always be able to do so. According to David Touretzky