Given how AI is the rage these days in just about every sphere of life, it would be natural to assume that the rather elaborately titled, How to AI: Cut through the hype. Master the basics. Transform your work, is yet another of those self-help and upskilling books by wannabe techie-turned-authors that claim to make you an AI wizard in a few hundred pages of jargon. Fortunately, it is not that. In fact, How to AI, as its author Christopher Mims says in the introduction, is “a book for people who want to use AI to get things done.” And it does so simply and sensibly.
Heard of AI? Need to use AI? Confused by AI? This book is for you
How to AI is basically an AI primer targeted at those who have heard of AI but get confused by it. It is a book designed to get general users up to speed with what AI is, what it can do, and how one can use it. It is not a textbook or a comprehensive guide, but rather a quick, easy read about the main parts of AI and how they are relevant to many of us. Mims is well-versed in tech, being a regular tech columnist in the Wall Street Journal and host of the Bold Names podcast, and does a good job. He steers clear of too much jargon and delivers what we think is one of the most accessible and easy-to-read books on AI — no specialised tech knowledge is needed to read and understand it. If you use a smartphone regularly, there is a fair chance that you will understand it.
Mims has arranged information in the form of what he calls AI Laws. These are not hard and fast rules cast in iron, but more like broad guidelines, and serve as milestones in this brief AI trip. He also intersperses the text with plenty of examples and does a lot of jargon-busting (LLM, Agentic AI, AGI, VLA models and all are explained) without ever getting too heavy. The book has about 250 pages in all, 25–30 of which cover the introduction, acknowledgements, notes and index, making this the kind of book one can easily read in a few days. Mims writes with a journalist’s easy flow, blending conversations and incidents with information, making this a very interesting read even for those comfortably conversant with AI.
‘AI is an assistant, not a replacement’, ‘AI is feature, not a product’, and more
This is not a book for those wanting to debate AI. Mims steers clear of trying to paint AI as an angel that will make our lives super easy or as a devil that will take over the world and deprive us of our livelihoods. His broad outlook is that AI is more of an able assistant than a creator — his very First Law of AI is “AI is an assistant, not a replacement” (it is the most important law of all, and one he reiterates at the end), while the Third Law reiterates the point: “AI is a feature, not a product.”
Mims time and again points out the pitfalls of trusting AI blindly, highlighting that it is a technology in progress and that while it is constantly improving, it is not close to perfect and is unlikely to be. For instance, his Sixteenth Law of AI states that “‘Classic’ predictive AI is brittle and breaks when big changes occur”, which is explained using the example of the insurance business, which depends heavily on predictive analysis. Similarly, another law points out that “AI isn’t creative, but it can help you be”, and “When successfully implemented, AI scales up rote knowledge work” is another. Many urge caution — one Law bluntly states, “Don’t trust it, and always verify its work,” and another reminds us that “‘Garbage in, garbage out’ still applies.”
Most Laws are explained with examples (some are simple enough to be explained and understood in little over a page), and sprinkled liberally throughout the book are a lot of interesting tidbits, ensuring it does not become a dreary read. One reads about how Clorox used AI to come up with the concept of a toilet bomb, which became one of its most successful products, and how a sluggish bureaucratic setup made Google lose out on being the leader in modern AI. There are also references to just about every big name in AI, from Jensen Huang to Sam Altman, often with some deft personal touches. He touches on trends and gives one a bit of a snapshot of what the future may have in store — “The field of human endeavour most transformed by generative AI is coding,” is his final Law.
The idea of using Laws to highlight key principles of AI is a good one, as it breaks the book into clear compartments. Twenty-four Laws might sound like a few too many to keep in mind, but these are sprinkled across a dozen chapters, ensuring that it never gets overwhelming. Experts might hem and haw at the relative absence of detail, and activists might point out that there is not much talk of AI’s environmental impact, but How to AI is more about making readers comfortable with AI and providing them with a starting point for further detailed reading. It is the one AI book all of us need to read even as we enter the age of AI.
Story continues below this ad
How to AI: Cut through the hype. Master the basics. Transform your work
Christopher Mims
Headline Press
254 pp
Rs 699
