Swaine’s World

Everything You Always Wanted to Know about AI but Were Afraid to Ask

It’s Book Week and I’m recommending an AI book.

Michael Swaine

5 min readJan 29, 2024

Like Swaine’s World? Scroll to the bottom of this post to tip the author or to subscribe.

Image of the Week

Just some books on my shelf.

BOOK WEEK

Artificial Intelligence: What Everyone Needs to Know

I chatted with Jerry Kaplan recently about AI, its history and current developments. I was embarrassed that I hadn’t read his book Artificial Intelligence: What Everyone Needs to Know, so I promptly bought and read it. I don’t do book reviews in these occasional Book Week posts; what I do is recommend books I think you should read, and tell you why I think that. In this post I’ll recommend that you read Jerry Kaplan’s Artificial Intelligence: What Everyone Needs to Know, and I’ll tell you why.

But first let me address this legitimate question: Why am I recommending a book on artificial intelligence that was published in 2016? Didn’t the AI world change at the end of 2022 with the release of ChatGPT? Wouldn’t an AI book that is eight years old be hopelessly out of date?

Well that’s three questions, but I’ll let you have them. To respond, I need to tell you about Jerry Kaplan.

Here’s one item from Wikipedia: “While at the University of Pennsylvania, Kaplan wrote the software for the first all-digital keyboard instrument, the Synergy, sold by Digital Keyboards, Inc. In 1980. The Synergy was used by Wendy Carlos to compose Digital Moonscapes.”

You’d think that accomplishment would bring him a bit of fame, but it has been completely overshadowed by his subsequent achievements: pioneering tablet computing, founding several successful startups including Go Corp. and Teknowledge, writing the natural language query system that became Symantec Q&A, developing Lotus Agenda, and a few other things. These days he’s keeping himself sharp lecturing on artificial intelligence many places, including at Stanford University. And writing books.

What I’m saying is, he knows something about Artificial intelligence. And what he had to say about AI in 2016 is definitely worth your time. Why?

Because the developments that led up to ChatGPT and the Generative AI programs that are shaking things up today didn’t occur overnight. The key technological insights and breakthroughs had already been made by 2016: what has happened in the last decade is the beginning of the realization of capabilities clearly discernible in the mid-2010s.

Discernible to someone like Jerry Kaplan, that is. As he says in his Preface, “After more than five decades of research, the field of AI is poised to transform the way we live, work, socialize, and even how we regard our place in the universe.” And that’s exactly what’s happening today. Kaplan tells us how it came about, and explores what it all means.

The book is organized in the form of 58 questions about AI, with Kaplan’s answers. The questions run the gamut:

Definitional questions: What is artificial intelligence? Where did the term artificial intelligence come from? What is “strong” vs “weak” AI? What is machine learning? What is natural language processing?

Practical questions: Which jobs are most and least at risk? Who’s going to benefit from this technological revolution? How can we minimize future risks?

Philosophical questions: Can a computer “think”? Can a computer be conscious? Will I ever be able to upload myself into a computer?

Where the philosophical questions are open to debate, he presents competing answers, so you can make up your own mind, but he doesn’t shy away from making his own views clear.

In 36 pages, he provides a succinct history of AI, describing the two main threads of AI research, symbolic reasoning and machine learning. He tells you which is better for which kinds of problems. And he explains why machine learning has largely won the battle: “[T]he combination of more powerful computers, access to large amounts of training data, and machine learning techniques conspired to crack the problem and to deliver systems of practical and commercial importance.”

It’s a short book, under 200 pages, so when he gets into the practical implications of Generative AI, he can’t go deeply into how it will affect every profession or walk of life. What he does instead is to pick one area, the law, and look at how AI will affect all aspects of the law.

For example, AI will make it easy for people to do for themselves a lot of activities they currently use lawyers for — contracts and wills and bankruptcies and activities that don’t involve actual disputes. Lawyers won’t like that, but they will like AI tools that help them do research — unless being able to get research done more quickly means fewer billable hours!

And how about writing laws? Legislatures write our laws (with the help of lawyers) and the courts try to figure out what the convoluted language that comes out of committee meetings and last-minute amendments actually means. But what if our laws were written in a formal language, even a computer language? What if that made the interpretation of the law simply a matter of running the program with the data pertinent to the case at hand? Then there are the questions about how laws should apply to AI systems. And these: Should an AI system be allowed to own property? Can an AI system commit a crime?

The questions Kaplan deals with are as pragmatic as this and as deeply philosophical as how advances in machine intelligence influence our understanding of what it mans to be intelligent.

Let me leave you with this quote from the book:

“One of the remarkable achievements of modern AI could be couched as a discovery in search of an explanation: how simply finding correlations between enough examples can yield insights and solve problems at a superhuman level, with no deeper understanding or causal knowledge about a domain. It raises the possibility that our human efforts at explanation are little more than convenient fictions, grand yet imperfect summaries of myriad correlations and facts beyond the capacity of the human mind to comprehend.”

Now that I’ve recommended the book, I’m going to tell you why you might not want to run out and buy it. Because the updated and retitled version, Generative Artificial Intelligence: What Everyone Needs to Know is available for pre-order as I write this. So consider this a recommendation of that.

BEFORE YOU GO…

The Pragmatic Bookshelf

If you buy programming books, you really should sign up for the Pragmatic Bookshelf newsletter. You’ll be among the first to hear about the newest releases, money-saving promotions, and upcoming events. And you’ll get 25% off your next purchase. They won’t sell or rent your email addresses to third parties, and you can unsubscribe at any time.

ICYMI

You can read all the 2023 back numbers of Swaine’s World at my blog home.

Coming Attractions

Next week I’m sharing one writing trick that I’ve found really useful.