AI Book Review: How To Think About AI – A Guide for the Perplexed by Richard Susskind

A concise guide to making sense of AI’s promises and threats, and a much-needed mental toolkit for leaders, lawyers and citizens trying to think clearly about it.

Article link: https://open.substack.com/pub/johanosteyn/p/ai-book-review-how-to-think-about?r=73gqa&utm_campaign=post&utm_medium=web&showWelcomeOnShare=true

Audio summary: https://youtu.be/iNcdelpDCLs

Follow me on LinkedIn: https://www.linkedin.com/in/johanosteyn/

I read a lot of books on artificial intelligence, partly because it underpins my work, but also because I am genuinely trying to understand the world my child will inherit. There is no shortage of material telling us that AI will either fix everything or end everything. What we lack are calm, structured ways of thinking about these claims. In How To Think About AI: A Guide for the Perplexed, Richard Susskind offers exactly that: not a breathless prediction of the next app, but a disciplined framework for anyone who feels overwhelmed by the noise and wants to think more clearly about what AI is, what it is not, and what it might become.

CONTEXT AND BACKGROUND

Susskind is hardly a newcomer to this conversation. Known globally for his work on law and technology, he has been engaging with AI since the early 1980s. In this book, published by Oxford University Press in 2025, he distils four decades of experience into a short, non-technical primer aimed at the intelligent general reader – with a particular eye on professionals, policymakers and, unsurprisingly, lawyers. He argues that weighing AI’s benefits and dangers has become one of the defining challenges of our age.

The book sets out to address confusion on several fronts: people are unsure what AI actually is, what current systems can and cannot do, how fast things may move, and whether we are looking at a new industrial revolution or an existential risk. It also tackles uncertainty about regulation and ethics – where to draw boundaries and who should decide. Running through it all is Susskind’s insistence that we are still at the very beginning of this story, with generative AI presented as a dramatic new chapter, but not the whole book.

INSIGHT AND ANALYSIS

How To Think About AI is not a technical manual. Instead, Susskind offers a set of lenses. He starts with a brisk history of AI, from early expert systems through winter periods to the deep learning surge and today’s foundation models. Along the way, he dismantles some familiar myths: that AI is a single technology, that it can already “think” as we do, or that progress is smooth and predictable. He is good at separating hype from genuine breakthroughs, and at explaining why something like ChatGPT can feel so impressive while still being built on pattern-matching rather than understanding.

A central contribution of the book is its focus on scenarios rather than forecasts. Susskind sketches a range of plausible futures, from largely beneficial augmentation of human capabilities through to more dystopian outcomes in which powerful systems undermine employment, privacy, democracy or even human survival. He does not claim to know which path we will follow. Instead, he offers readers questions they should be asking: about control, alignment, incentives and governance. This is especially helpful for leaders who may never read a technical paper but must still make strategic decisions about AI adoption and risk.

Unsurprisingly, there is a strong legal thread. Susskind spends time on how AI will reshape professions, courts, contracts and regulation, and on the distinction between focusing on processes and focusing on outcomes. For South African lawyers and policymakers, these sections will feel very relevant. He challenges the idea that our task is simply to map existing rules onto new tools. Instead, he suggests that AI forces us to revisit what we mean by concepts like responsibility, explanation and fairness in systems where outcomes emerge from opaque statistical processes.

IMPLICATIONS

For business and government leaders, the book functions as a practical thinking aid. It will not tell you which vendor to choose or which use case to prioritise, but it will help you ask better questions. Susskind pushes readers to think in terms of systems: who designs and trains the models, whose data they ingest, who bears the risk when things go wrong, and how power might shift as AI moves deeper into decision-making. In a South African context, where inequality and institutional fragility are already pressing concerns, that systems perspective is crucial.

For lawyers, regulators and academics, How To Think About AI is almost required reading. It situates current debates about liability, transparency and regulation in a broader narrative about technological change, reminding us that our legal frameworks were built for a very different world. At the same time, it resists fatalism. Susskind argues that law and policy can still play a meaningful role in steering AI, provided we do not wait until the most powerful systems are entrenched.

For ordinary citizens and parents, this is also a useful book. It does not drown the reader in jargon, and it does not patronise them either. Instead, it offers a way to think through the questions that keep many people awake at night: will there be work for my children, who will control these systems, and how do we stop AI from being used primarily to entrench existing privilege?

CLOSING TAKEAWAY

How To Think About AI: A Guide for the Perplexed is precisely what its title promises: a guide to thinking, not a handbook of answers. Richard Susskind brings long experience, clear writing and a healthy scepticism to a noisy debate. Readers looking for breathless predictions or easy reassurance will not find them here. Instead, they will encounter a careful mapping of possibilities, risks and responsibilities.

In my view, this makes the book particularly valuable for South African and African readers who must make hard choices about scarce resources, fragile institutions and a rapidly changing global landscape. We may disagree with some of Susskind’s assumptions or emphases, but the discipline he brings to the question – how should we think about AI at all – is exactly what our public conversation needs.

Author Bio: Johan Steyn is a prominent AI thought leader, speaker, and author with a deep understanding of artificial intelligence’s impact on business and society. He is passionate about ethical AI development and its role in shaping a better future. Find out more about Johan’s work at https://www.aiforbusiness.net