Artificial Intelligence Complete: Fundamentals and Applications

By Fouad Sabry

What Is Artificial Intelligence Complete

In the field of artificial intelligence, the problems that are considered to be the most challenging are informally referred to as AI-complete or AI-hard. These names imply that the difficulty of these computational problems, presuming that intelligence is a function of computation, is equivalent to that of solving the central problem in artificial intelligence, which is the process of making computers as intelligent as people, also known as strong AI. When applied to an issue, the term "AI-complete" indicates that the problem cannot be solved using a straightforward algorithmic approach.

How You Will Benefit

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

Chapter 1: AI-complete

Chapter 2: Vehicular automation

Chapter 3: Artificial general intelligence

Chapter 4: Computer vision

Chapter 5: Automated theorem proving

Chapter 6: Natural-language understanding

Chapter 7: Automated reasoning

Chapter 8: Synthetic intelligence

Chapter 9: Adaptive system

Chapter 10: Bongard problem

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

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

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

• Language: English
• Publisher: One Billion Knowledgeable
• Release date: Jul 4, 2023

Book preview

Chapter 1: AI-complete

In the field of artificial intelligence, the problems that are considered to be the most challenging are informally referred to as AI-complete or AI-hard. These names imply that the difficulty of these computational problems, presuming that intelligence is a function of computation, is equivalent to that of solving the central problem in artificial intelligence, which is the process of making computers as intelligent as people, also known as strong AI. When an issue is described as being AI-complete, it indicates that its solution cannot be reduced to a simple, deterministic method.

It is hypothesized that AI-complete issues include computer vision, natural language comprehension, and the ability to cope with unforeseen conditions while addressing any real-world task.

Fanya Montalvo is credited with the invention of the word, which is an analogy with NP-complete and NP-hard in complexity theory. This theory officially characterizes the most well-known class of challenging tasks.

It is suggested that AI-complete issues consist of the following::

AI peer review (composite natural language understanding, automated reasoning, automated theorem proving, formalized logic expert system)

Bongard problems

The use of computer vision (and subproblems such as object recognition)

Natural language comprehension (and subproblems such as text mining,)

Autonomous driving

coping with unforeseen events while finding solutions to problems arising in the real world, whether it be navigation, planning, or even the type of reasoning that is carried out by expert systems.

It is necessary for a machine to comprehend the text in order for it to provide an appropriate translation. Therefore, it must possess at least some capacity for reasoning in order to comprehend the author's line of reasoning and follow it. In order for it to understand what is being spoken about, it must have a comprehensive understanding of the world; at the very least, it must be acquainted with all of the same commonplace facts that a typical human translator is aware of. Some of this information is in the form of facts that can be explicitly represented, while other aspects of this knowledge are subconscious and closely associated with the human body. For instance, the machine may need to understand how the sensation of being near an ocean makes one feel in order to accurately translate a particular metaphor that is found in the text. It is also necessary for it to model the aims, intents, and emotional states of the writers so that they may be correctly reproduced in a new language. In a nutshell, the machine has to be able to reason, have commonsense knowledge and the intuitions that underpin motion and manipulation, have perception, and have social intelligence. These are some of the essential human intellectual talents. As a result, it is generally accepted that machine translation has reached the level of artificial intelligence required to perform at the same level as humans.

The currently available AI systems are only able to tackle very simplified and/or constrained versions of AI-complete issues; they cannot answer these problems in their entire generality. When researchers in artificial intelligence (AI) attempt to scale up their systems to handle more complex, real-world situations, the programs tend to become excessively brittle without having commonsense knowledge or a fundamental understanding of the situation. They fail when unexpected circumstances that are outside of the context of the original problem begin to appear. When humans are confronted with novel circumstances in the world, they are afforded a great deal of assistance by the fact that they are well-versed in what to anticipate: they are aware of everything in their immediate surroundings, including what those surroundings are, why they are there, and what they are likely to do. They are able to notice odd circumstances and change themselves appropriately. A computer that lacks robust artificial intelligence has no other capabilities to rely on.

The study of the relative difficulty of computing different types of computable functions is what computational complexity theory is all about. It does not include issues for which the solution is unknown or which have not been officially characterized. This is because its definition precludes such coverage. The traditional theory of complexity does not provide the definition of AI-completeness since many AI issues do not yet have formalizations.

A complexity theory artificial intelligence has been presented as a solution to this issue. It is based on a paradigm of computing that divides the burden of computation between a person and a computer: one portion of the problem is handled by the computer, while the other half of the problem is solved by the human. This is formalized via the use of a Turing machine with human assistance. The formalization establishes algorithm complexity, issue complexity, and reducibility, which in turn makes it possible to establish equivalence classes.

The complexity of executing an algorithm with a human-assisted Turing machine is given by a pair \langle\Phi_{H},\Phi_{M}\rangle , where the first component stands for the difficulty of the role played by the human and the second component stands for the difficulty of the part played by the machine.

The difficulty of addressing each of the questions listed below using a Turing computer in conjunction with a human is as follows::

Optical character recognition for printed text: \langle O(1), poly(n) \rangle

Turing test:

for an n -sentence conversation where the oracle remembers the conversation history (persistent oracle): \langle O(n), O(n) \rangle

for an n -sentence conversation where the conversation history must be retransmitted: \langle O(n), O(n^2) \rangle

for an n -sentence conversation where the conversation history must be retransmitted and the person takes linear time to read the query: \langle O(n^2), O(n^2) \rangle

ESP game: \langle O(n), O(n) \rangle

Image labelling (based on the Arthur–Merlin protocol): \langle O(n), O(n) \rangle

Image classification: human only: \langle O(n), O(n) \rangle , and with less reliance on the human: \langle O(\log n), O(n \log n) \rangle .

{End Chapter 1}

Chapter 2: Vehicular automation

Assisting a driver via the employment of technologies like as mechatronics, artificial intelligence, and multi-agent systems is what is meant by the term vehicular automation (car, aircraft, watercraft, or otherwise).

A fully automated driving system is often an integrated package consisting of a number of separate automated driving technologies that work together. When driving is automated, it means that the driver has given the power to drive (including all relevant