The 5 Best Books About AI and the Future of Work

If The Last Skill is the book for the 2 a.m. existential spiral, Co-Intelligence is the book for 9 a.m. Monday. Ethan Mollick is a Wharton professor who has logged more hours collaborating with AI systems than almost any academic I know of, and the book reads like it — full of real experiments, real surprises, and real failures.

His core move is reframing AI from “tool” to “collaborator.” That sounds like semantics, but it changes everything about how you approach the technology. A tool you pick up and put down. A collaborator you negotiate with, push back on, learn from. Mollick gives you concrete ways to do that across writing, analysis, brainstorming, teaching, and hiring.

The best parts of this book are the ones where Mollick admits he was wrong about something. There’s a section where he describes an AI experiment that produced results so good he initially refused to believe them — then had to redesign his entire course. That kind of honesty is rare in a field full of evangelists.

The gap: Mollick writes from inside a business school. The book is strongest on knowledge work and weakest on labor that doesn’t happen behind a laptop. If you’re a nurse, a carpenter, or a line cook, you’ll find his framework useful but may have to do translation work on your own.

Read this if: you need to work with AI starting tomorrow and want practical, tested advice from someone who has actually done it — not just theorized about it.

The case for

Hoffman argues that AI is an amplifier, not a replacement. His version of the future of work isn’t mass unemployment — it’s mass empowerment. Individual professionals gaining capabilities that previously required entire teams. Small businesses competing with enterprises. Founders in Lagos and Lima building products that rival those from San Francisco. He calls this “superagency,” and when he’s at his best, the vision is genuinely compelling.

The asterisk

Reid Hoffman is one of the largest individual investors in AI. He co-founded LinkedIn, sits on the board of Microsoft, and has backed multiple AI startups. He is, to put it plainly, a man who profits enormously when you feel good about AI. The book acknowledges this, but acknowledgment and correction are different things. There are chapters that read more like investment theses than career advice.

What saves it

Beneath the optimism, there’s serious thinking here. Hoffman’s chapter on how AI changes the economics of expertise — making rare knowledge common and common knowledge worthless — is one of the sharpest analyses I’ve read anywhere. And his framework for thinking about which tasks get automated versus which get amplified is genuinely useful, even if you disagree with where he draws the lines.

Read this if: you’ve been marinating in doom and need a rigorous counterargument, or if you’re an entrepreneur trying to figure out where the real opportunities are.

This is the most structured book on the list, and that’s both its strength and its limitation. Bornet spent years at McKinsey and brings that consulting DNA to everything — frameworks, matrices, three-part models. His central argument boils down to three skills he believes will remain uniquely human: genuine creativity (not recombination), critical thinking (not pattern matching), and social authenticity (not performance).

I’ll be honest: parts of Irreplaceable feel like a very good corporate workshop. The language sometimes drifts toward the boardroom. “Leverage your uniquely human competencies to create differentiated value” — that kind of thing. If you’ve spent time in management consulting, you’ll feel right at home. If you haven’t, you might need a translator.

But here’s why it’s on this list: Bornet does something few other authors attempt. He gives you a concrete self-assessment. Not a personality quiz, but a real diagnostic framework for evaluating which parts of your current work are at risk and which parts aren’t. For anyone in a corporate role trying to make a strategic case — to their boss, their team, or themselves — about why their job still matters, this book hands you the vocabulary and the evidence.

Read this if: you work in a corporate environment, think in frameworks, and need a structured way to evaluate your career exposure to AI — especially if you’re the person who has to present that analysis to leadership.

The oldest book on this list, and in some ways the most important.

Stuart Russell co-wrote the textbook on artificial intelligence — the one used in virtually every university AI course on the planet. When he says the standard approach to building AI is fundamentally broken, it carries a weight that most critics can’t match. His argument is precise: we’ve been building AI systems that optimize for objectives we specify, when we should be building systems that defer to objectives we hold. The difference sounds subtle. It isn’t.

Why does a safety book belong on a “future of work” list? Because Russell makes you see something the other four books don’t address directly: the future of work depends entirely on whether we build AI that respects human authority. If we get alignment right, AI becomes the collaborator Mollick describes and the amplifier Hoffman envisions. If we get it wrong, the career advice doesn’t matter because the systems won’t be listening to us anyway.

Published in 2019, some of the technical specifics have aged. The conversational AI examples feel quaint compared to what we have now. But Russell’s core framework — that beneficial AI must be uncertain about human preferences and willing to be corrected — has only become more relevant as the systems have grown more powerful. The fact that this book predates GPT-3, yet accurately describes the alignment problems we’re struggling with in 2026, tells you something about how clearly Russell was thinking.

Read this if: you want to understand the structural question underneath all the career advice — whether the AI systems we’re building will actually work with human values or just work around them.