Artificial Intelligence: A Very Short Introduction

Very Short Introductions #575

by Margaret A. Boden

The applications of Artificial Intelligence lie all around us; in our homes, schools and offices, in our cinemas, in art galleries and -- not least -- on the Internet. The results of Artificial Intelligence have been invaluable to biologists, psychologists, and linguists in helping to
understand the processes of memory, learning, and language from a fresh angle.

As a concept, Artificial Intelligence has fueled and sharpened the philosophical debates concerning the nature of the mind, intelligence, and the uniqueness of human beings. In this Very Short Introduction, Margaret A. Boden reviews the philosophical and technological challenges raised by Artificial
Intelligence, considering whether programs could ever be really intelligent, creative, or even conscious, and shows how the pursuit of Artificial Intelligence has helped us to appreciate how human and animal minds are possible.

ABOUT THE SERIES: The Very Short Introductions series from Oxford University Press contains hundreds of titles in almost every subject area. These pocket-sized books are the perfect way to get ahead in a new subject quickly. Our expert authors combine facts, analysis, perspective, new ideas, and
enthusiasm to make interesting and challenging topics highly readable.

Genres: Science, Nonfiction, Artificial Intelligence, Philosophy, Technology, Computer Science, Education

164 pages, Paperback

First published January 1, 2018

Reviews

[Martin]

I good overview of the history of AI, although I was expecting the book to be more of an explanation of it.
Some of the concepts like phenomenal consciousness, rational and relevant thinking, and also social and emotional intelligence would be interesting to follow up on in the "Further reading" section.

[David A Townsend · Rating 4 out of 5]

Logic gives way to statistics, single units to collective, and deterministic purity to probabilistic noise.

[Gerhard · Rating 3 out of 5]

'Moreover, AI has focused on intellectual rationality while ignoring social/emotional intelligence—never mind wisdom. An AGI that could interact fully with our world would need those capacities, too. Add the prodigious richness of human minds, and the need for good psychological/computational theories about how they work, and the prospects for human-level AGI look dim.'

Okayish as a basic historic overview, but incredibly dated in terms of the latest developments and thinking. Needs a complete overhaul.

[Nick Davies · Rating 4 out of 5]

Fascinating. I've been aware of these 'A Very Short Introduction...' books for a few years, friends and family having found others in the series worthwhile and interesting - but I had not myself read one until now. Artificial Intelligence is a subject about which I am curious but do not have a particularly full or good understanding, and this proved to be a very good introduction to, and stimulus of further desire to find out more about, the subject.

My only criticism, which I acknowledge is an obvious consequence of the format, was that it was at times hard to follow. Compressing a large and complex field into a short and readable format was always necessarily going to lead to a book I think I'll need to re-read a couple more times to ensure I absorb and understand it more fully.

[sulaiman al-maawali · Rating 4 out of 5]

يبدو من الواضح أن تقدم الذكاء الاصطناعي مرتبط بتقدم العلوم الأخرى مثل النفس والأعصاب والمنطق والرياضيات واللغة والأحياء وغيرها.

هذه المقدمة تطرح نقاشات وأسئلة مثل ما إذا كان الذكاء الاصطناعي قادرا على فهم الكتابة والصور مثل ما يفهم البشر، بمعنى هل أنظمة الذكاء الاصطناعي قادرة على فهم حقيقي بحيث يمكن أن يكون لها إدراك حسي وبصري واع؟ وهل يمكن وجود ما نسميه ذكاء وعقل في الآلة يوازي ما هو موجود لدى البشر (أو ربما حتى الحيوان) بدون وجود حياة تصاحب هذا الذكاء؟ وإلى أي مدى يمكن أن يكون الذكاء الاصطناعي قادرا على محاكاة الدماغ؟ ما هو المنظور الذي يرى منه الذكاء الاصطناعي وعلى أي أساس يتم تحديده؟ وهل طبيعة الذكاء الاصطناعي تحاول الوصول نفس طريقة تفكير الذكاء البشري؟

أعجبني المثال المطروح حول ما اذا كان الذكاء الاصطناعي له القدرة حاليا أن يفكر بنفس طريقة البشر:
"اذا كان الرجل بعمر ٢٠ يجمع عشرة أرطال من التوت الأسود في ساعة، والمرأة في عمر ١٨ عامًا تجمع ثمانية أرطال في الساعة فكم عدد الأرطال التي سيجمعونها إذا تعاونا معا؟
بالتأكيد الإجابة ليست -بالضرورة- ١٨ رطلا. قد تكون الكمية أكبر بكثير (لأن كليهما سيستعرض مهاراته)، أو الأرجح أنها ستكون أقل.
ما أنواع المعرفة المتضمنة في هذا المثال؟ وهل يتمكن الذكاء الاصطناعي العام من التغلب على ما يبدو أنه معطيات حسابات رياضية بسيطة؟"
بعد تجربة طرح السؤال أعلاه ل ChatGPT، هذه هي الاجابة التي تم الحصول عليها والتي تؤيد ما ذكره الكاتب، أي أنه "قد تكون الكمية أكبر بكثير (لأن كليهما سيستعرض مهاراته)، أو الأرجح أنها ستكون أقل"، لكن رد ChatGPT في نسخته الحالية مبني على الحسابات الرياضية فقط كما يتضح في هذه الإجابة:
"إذا تعاون الرجل والمرأة معًا في جمع التوت الأسود، فإنهما سيجمعان معًا عدد الأرطال التالي:
‎عدد الأرطال التي يجمعها الرجل في ساعة واحدة = 10
‎عدد الأرطال التي تجمعها المرأة في ساعة واحدة = 8
‎إذاً، يمكن حساب عدد الأرطال التي يجمعونها معًا في ساعة واحدة عن طريق جمع عدد الأرطال التي يجمعها الرجل والمرأة:
10 + 8 = 18
‎وبالتالي، سيقوم الرجل والمرأة معًا بجمع 18 رطلاً من التوت الأسود في الساعة الواحدة."

يمكن القول أن الكتاب بشكل عام رائع ويطرح العديد من التساؤلات بالرغم من وجود ملاحظات في الترجمة في النسخة العربية.
ملخص فصول الكتاب باللغة الانجليزية يمكن الاطلاع عليها من خلال: https://academic.oup.com/book/415

[David · Rating 2 out of 5]

A challenging read for the novice in AI. The author is deeply knowledgeable about this and related subjects, but the book suffers from overly dense and detailed prose that is not approachable by many. A careful read will illuminate all of the ideas in this comprehensive work and will be useful to many who make it through to the end, but it will likely be slow going.

[Craig French · Rating 1 out of 5]

Some good references and good for pointers for where to follow up themes. But dreadful as an introduction - lost count of how many key ideas, concepts, and notions were unexplained. No very clear narrative to the book either. As a starter for AI look elsewhere.

[Rick Sam · Rating 5 out of 5]

An Excellent Introduction to Artificial Intelligence.

Here is the outline:

1. What is Artificial Intelligence?
2. General intelligence as the Holy Grail Language,
3. Creativity, emotion Artificial neural networks
4. Robots and artificial life
5. But is it intelligence, really?
6. The Singularity

The Field of A.I is moving fast, researchers pouring volumes of research. Researchers are from all backgrounds in the field of A.I, it's interdisciplinary -- Psychology, Philosophy, Computer Science, Mathematics

In SaaS World, I could see many parts of this field, actualized into wonderful products solving real-world problems.

The Goal for field in A.I is to move towards General Intelligence, which we are far, far away.

I would recommend this to anyone, who wants to know about A.I.

Deus Vult,
Gottfried

[Wing · Rating 3 out of 5]

So-called Artificial Intelligence does not merely mean the framing of objectives and achieving them through the search for effective sequences of means via symbolic logic. It is also about the tolerance of ambiguity and incompleteness by the system through feedback and remodelling in order to self-organise and generate probabilistic inferences. The aim to is discover optima and not imaginary perfections. The challenge has always been to be domain-general and open-ended and yet remain contextual and relevant. The field of robotics highlights the importance of incorporating situated responses into a representational context, and expanding the possibility space at the same time. The brief chapter on consciousness, while introductory in depth is comprehensive in scope. In sum, that psychological processes can be modelled by machines does not mean the brain itself is actually a machine. The main lesson to learned from AI is that human minds are far more richer and subtler than psychologists previously imagined. Overall the book suffers from being disjointed and fragmentary, and is marred by the appearance of terms and jargons without any sufficient explanation to the reader. However the topics covered are quite fascinating. Three stars.

[Eric Carzon · Rating 4 out of 5]

You ever read a book you only understood about 1/3 of? This is that book. But the 1/3 or so I feel I understood was very illuminating. In a very compact format, the author describes the history of Artificial Intelligence (Ai), which includes the interesting evolution of some major rifts and faultlines between some of the major players that have significance today. Many concepts are explored, and I got a further appreciation of Alan Turing, who had a larger influence on Ai thinking that I realized. I highly recommend this book.

[Ramona · Rating 3 out of 5]

Interessant, maar vond het wel heel “dense”.

[Pete · Rating 4 out of 5]

Artificial Intelligence : A Very Short Introduction by Margaret Boden is an interesting short book about AI. Boden is a professor of Cognitive Science who has been writing books about AI for over 40 years.

The book tends toward the historical and details the fairly short history of AI.

Boden starts by discussing what we mean by AI. General Intelligence is then descibred. The work of people in Good Old Fashioned AI (GOFAI) in the 60s and 70s by people like Marvin Minsky is written about. Then the connectivist triumph with artificial neural nets and back propagation in the 1980s and beyond is described. Boden goes on to write about Robots and AI and discusses if what has been achieved is really intelligence. Finally a chapter on The Singularity is written up.

Boden comfortably goes through all the major moments, but there is little description given to what many of the things are and their applications, which might be expected.

Just to add, here is something from one of the latest tools in AI, Chat GPT on The Very Short Introduction Series:

Very short introduction books offer a concise and comprehensive overview of a particular subject, making them a great resource for those who are looking to gain a broad understanding of a topic in a short amount of time. These books are designed to provide a quick but thorough introduction to a subject, covering its history, key concepts, and current state of knowledge in an accessible and engaging manner. They are ideal for anyone who wants to gain a solid foundation in a new subject, whether it be for personal or professional development. Additionally, they are perfect for busy individuals who have limited time to devote to reading, as they can be read and absorbed quickly, providing a convenient and effective way to learn about a subject without having to invest a lot of time and effort.

The book’s final chapter on the Singularity has a discussion of where AI might displace people, such as in some office work including writing summaries. Boden is skeptical towards Kurzweil and other’s claims of the Singularity.

Artificial Intelligence : A Very Short Introduction is worth reading for anyone interested in AI. It’s not one of the excellent introduction to AI courses available online such as Andrew Ng’s excellent course, but it is a worthwhile book.

[Kyle · Rating 2 out of 5]

It seems like the most prescient part of understanding the potential of artificial general intelligence is knowing the many, many acronyms and abbreviations. Perhaps if the programmers and researchers spent a little less time lining up letters to christen each virtual machine, we'd probably have digital match for the human mind. Singularity-skeptics like Boden may not like this assumption, and she seems a bit too wrapped up in disproving in the last chapter what had been proposed in the preceding six, but nevertheless knows her way around this history of acronyms to say something meaningful about human-computer relations. Unless that's exactly what the Matrix wants people to believe.

[Allen Martin · Rating 3 out of 5]

I’m giving this 3 stars not because of the content (which is an excellent introduction) but because of the writing style. It seems like some key concepts or acronyms are used, but they aren’t defined until later. Sound I found myself flipping back and forth through the book

I much prefer Melanie Mitchell’s book as an introductory work.

[Viviana Letizia · Rating 5 out of 5]

Fan of the series A very short introduction, this one is a must read for the current time.

[Daniel · Rating 4 out of 5]

This book may be straining the definition of "Introduction," as in a good first book to read on a subject. It covers an extraordinarily broad and complex technical field, with nearly every page and often paragraph mentioning yet another concept, software system, or thinker about which entire books could be (and often have been) written. Certainly Wikipedia articles have been written for most of them, and I spent about as much time looking them up as reading about them in the book. In most cases Wikipedia goes into more depth and it's usually needed. The book mostly made sense to me, but I've been reading about artificial intelligence occasionally since the 1980s.

The Very Short Introductions series divides other complex topics into subtopics (see for example the VSIs on astronomy, which separately cover stars, planets, galaxies, moons, astrophysics, cosmology, telescopes, and history). AI seems more than complex enough to justify a similar division. Then the reader might actually get introduced to a particular area of AI (symbolic, connectionist, deep learning, the Singularity, etc.) as in grasping how it works.

AI is a branch of computer programming, and it's hard to "understand" computer programming without actually programming computers (try to imagine learning mathematics without working math problems). Thus there are probably limits to how much of AI can be conveyed merely by reading books. Ideally, an introduction to the field (short or otherwise) should include actual programs one can run and study. But the author makes a brave attempt to convey the wonders of AI through prose only, and one hopes the interested reader will follow up with the Further Reading recommendations. Or at least look up all book topics on Wikipedia.

[Gerges Badr · Rating 5 out of 5]

غالبًا ما يهدف الذكاء الاصطناعي إلى تقديم أدوات مفيدة ولا يهدف إلى محاكاة ذكاء الإنسان. بل إلى جعل المستخدمين يعتقدون أنهم يتعاملون مع إنسان.

هو فعلًا مقدمة قصيرة جدًا، تناول الذكاء الاصطناعي من أبعاد مختلفة ومتنوعة لتكوين نظرة عامة أكتر عن الموضوع، وبشكل جيد جدًا. عرّف الكتاب الذكاء الاصطناعي وأنواعه وإزاي بيشتغل من بداياته وصولًا لوقتنا الحالي، وفي خلال الكتاب كله وفي أقسام مختلفة بتناقش المؤلفة حاجات ملهمة جدًا في وجهة نظري، زي علاقته بالحواس الإنسانية وهل هو قادر على محاكاتها فعلًا؟ أو دور الأخلاق في الذكاء الاصطناعي، أو هل يمكن للذكاء الاصطناعي تحمل المسؤولية؟ ولو حدث خطأ ما بسببه فمن سيتحمل تلك المسؤولية؟ هل الذكاء الاصطناعي قادر على "فهم" أشياء مثل اللغة أو الإبداع أو العاطفة وبأي شكل وإلى أي مدى؟ وفي فصل كامل - السادس - تطرح مسألة ما إن كان الذكاء الاصطناعي "ذكي حقًا"!

أهم وأفضل فصل بالنسبة لي كان الفصل السابع (التفرد)، والتفرد يعني وقتًا معينًا من الزمن تصبح فيه الآلة أذكى من الإنسان، هل هذا ممكن حتى؟ وبأي شكل؟ ومتى؟ وما مقدار المخاطرة المحتملة إن حدث ذلك؟ وهل يمكننا كجنس بشري تجنبها؟ بالطبع كل تلك المناقشات (بين مؤيد ومعارض) بشكل علمي ومنهجي وواقعي بعيدًا عن دراما هوليوود.

مشكلة الكتاب الوحيدة هي أن المؤلفة وبرغم كونها عالمة وباحثة خبيرة وكبيرة إذ يتضح هذا من معلوماتها وطرحها للموضوع عمومًا، إلا أنها ليست "كاتبة" في المقام الأول من وجهة نظري، كان هذا واضحًا في الفصل الأول تحديدًا - الممل - وفي بعض الاقسام القليلة الأخرى على مدار الكتاب، مصطلحات غريبة وطريقة صعبة في الشرح على أي حد مش خبير في الذكاء الاصطناعي.
لكن إجمالًا الكتاب رائع ويُنصح به بشدة لكل حد مهتم بالموضوع.

كانت أولى قراءاتي في السلسلة، وأول خطوة في هدف إنهاء أكبر عدد ممكن من الكتب منها.

[Paul · Rating 5 out of 5]

Many advances in artificial intelligence have been made over the past several years, starting with Siri and Alexa. This book gives the details.

The science of AI involves many other sciences, including neurophysiology, logic and psychology. A major problem is how to present a problem, or a question, to a computer in a way that the computer will understand. Another problem is how to show things like emotion and creativity. What is consciousness? Is intelligence more than just IQ or the Turing Test?

Robot designs have had better luck creating robots that resemble insects, with six or eight legs, than in creating robots that look human, with two legs. When is The Singularity coming? When is Skynet coming? The general answer from this book is: not anytime soon.

This is a very well done introduction to the world of artificial intelligence. Some of it gets rather technical, but most of it is good for the general reader. It is very much recommended for anyone who wants an introduction to AI. Veterans of the field might also learn a thing or two.

[Hank Hoeft · Rating 4 out of 5]

Some volumes in the Very Short Introduction series are like primers, delivering basic information to readers who are not already familiar with the topic, while others are overviews, brief summaries that cover the entire breadth of the subject—overviews that due to their brevity, can be difficult for readers not already somewhat familiar with the subject. Artificial Intelligence: A Very Short Introduction is one of the latter sort, so dense with information that I had to read it slowly in order to digest it, and even then I didn’t completely understood all of it. Still, if one already possesses the schema necessary to follow along, this is an excellent overview of the subject, if not the easiest "introduction."

[James Stuart · Rating 4 out of 5]

Very detailed historical references show that AI is an expansive field with deeper roots than just Deep Learning! But there's also an excessive academicism in the ever-present skepticism over recent results and perspectives for near-term AGI. So the book's predictions, from 2018, might not age so well. Already in 2021, many of the current results in NLP and autonomous cars seem to put into question some of the author's reservations about the generality of current approaches.

Extremely useful for broadening your perspective on the history and the lesser-known areas of AI. To be combined with more optimistic (albeit uncertain) perspectives from authors such as Kurzweil or Tegmark.

[Sasha · Rating 5 out of 5]

The book "Artificial Intelligence: A Very Short Introduction" is a comprehensive guide to the fundamental concepts and applications of AI. While the book is not simple and requires a basic familiarity with AI terminology, it covers all the essential areas of AI that one should know. The author provides a clear and concise explanation of AI's history, techniques, and contemporary issues, making it an excellent resource for students pursuing studies related to AI. Overall, this book is a valuable read for anyone seeking to understand the basics of AI and determine which area of knowledge to pursue.

[Sumir Sharma · Rating 5 out of 5]

Read it early as it has an expiry date

Margaret has given history of development of AGI in a narrative manner. I would is written as I'd. It may not be liked by some readers. I am taken over by Chapter 6. Chapter 7 talks about future of
ASI replacing AGI and important debates. It is pertinent to note here that the author has been an active participant in development. On the sideline, she is witty also.

[Marc Chéhab · Rating 2 out of 5]

I read the first chapter, skim-read the second, glanced over the rest... The books problem IMO is that it simply enumerates all things that exist in the AI space without actually explaining pretty much any of it.

If you want to understand what AI does, there's an epic intro by Bert Hubert here: https://berthub.eu/articles/posts/hel... Also, I recommend Wolfram's "What is ChatGPT doing and why does it work?"

[Ralph Palm · Rating 2 out of 5]

a nigh-useless catalogue of career autobiography and score settling. one wonders if an LLM could have written a better version of this. there are occasionally informative elements, but they seem mostly accidental. The quality of the Oxford Very Short Introduction series varies considerably, and this seems like a good example of its nadir.

if you detect any hostility in my tone, it's because I am annoyed that this book wasted my time.

[Cem Yüksel · Rating 3 out of 5]

Quite condensed introduction to AI from early days to the present. This content creates a wave of theories by different people which are not explained well due to short introduction giving a taste of “not completed”. The chapter on consciousness is a good one for different discussions on AI’s future and discussion between the AI’s limits .

[Clo Willaerts · Rating 3 out of 5]

Handy if you’re looking for clear definitions

This book is an ideal starting point if you’re writing a paper or article about AI, and need to quote clear definitions of terms like Machine Learning or Deep Learning.
It is indeed very short and ends abruptly without making a clear point - besides maybe that General Artificial Intelligence is still distant.

[Joe.grimbook · Rating 2 out of 5]

Secondo il titolo originale dovrebbe trattarsi di un'introduzione alla materia, ma il testo fa riferimento a un linguaggio specialistico decisamente al di là delle capacità del lettore medio. Inoltre, molti concetti estremamente complessi vengono citati senza l'ausilio di alcun tipo di nota o inciso atto a spiegarli, vengono considerati automaticamente come già conosciuti dal lettore.

[Ryan]

This introduction covers so much research.

Currently, AI is limited.
The singularity/existential threats are unlikely, since ASI requires AGI first, and AGI may not be possible.
We should consider more what AI will do to everyone's jobs in the near future.

Also, an individual's use of current AI technology is powerful and can have risks.

[Bisman · Rating 3 out of 5]

Would NOT recommend if you are looking to quick a succinct summary of Artificial Intelligence. This book is just a long list of acronyms of random things related to AI. I was lost 4/5 of the book and didn’t find it engaging nor interesting. There are far better books on AI out there. I usually love the short introduction series but this one was far too chaotic and illogical.

[Marco Leonesio · Rating 3 out of 5]

Panoramica completa ma senza approfondimenti sull'evoluzione dei vari modelli di intelligenza artificiale. Non risulta chiaro se l'autrice ritiene che i limiti dell'AI siano legati alle implementazioni attuali (un fatto incontrovertibile), oppure se esistono veri e propri limiti teorici invalicabili. In quest'ultimo caso, questi non vengono spiegati chiaramente.