We’re Reading About AI Like Never Before, and That Matters
I stayed up until 3 AM last week reading about algorithmic bias and I texted nobody because what was I going to say? “Hey, did you know that five companies control over 80% of the computing power behind the large language models we use every day?” Not exactly midnight-crisis material, but it absolutely should be. The thing is, this isn’t obscure academic work anymore. The conversation about artificial intelligence and what it means for society has moved from think tanks into bookstores, and the books being published right now are doing something genuinely important: they’re making the incomprehensible legible.
Oxford University Press tracked something wild between 2023 and 2025. AI-themed nonfiction submissions jumped by 67 percent. That’s not just a blip. That’s publishers watching public anxiety spike and authors rushing to fill a void. The Frankfurt Book Fair confirmed it last year, identifying AI and society as the single fastest-growing nonfiction category with over 400 titles displayed from major publishers. When you walk into those conversations as a reader, you’re not arriving late. You’re arriving at exactly the right moment.
Mustafa Suleyman’s ‘The Coming Wave’ Got a Second Life, and It’s Worth Reconsidering
Here’s what I love about reading timing. ‘The Coming Wave’ dropped in 2023 to decent reception, but when Suleyman became CEO of Microsoft AI in 2024, people suddenly wanted to reread it with fresh eyes. The book climbed recommendation lists in 2025 not because the writing changed but because context changed everything. This is what attentive reading gives you: the chance to hear a work differently when circumstances shift around it.
The book’s central argument is straightforward but chilling. Suleyman contends that we’re approaching a technological inflection point where AI systems will be too powerful to control and too cheap to contain. He writes about this with the clarity of someone who has built these systems, which means he’s not theorizing from the outside. He’s describing something from lived experience. Whether you believe his solutions work is secondary to the real value here: he’s thinking carefully about power, and he’s thinking at scale. The sentences aren’t ornate or difficult. That’s precisely what makes them stick with you at 2 AM.
Kate Crawford’s Updated Essays Show That Power Structures Haven’t Changed, Only the Stakes
Kate Crawford published updated essays through MIT Press AI and Society catalog 2025 that expanded on the framework she built in ‘Atlas of AI’. What struck me most wasn’t the new data points, though they’re devastating. It was her analytical voice, which has sharpened since her earlier work. She writes with the precision of someone who has spent years watching how power actually operates inside technology companies.
The specific number matters here: five companies control over 80 percent of large language model compute capacity globally. That’s not a problem waiting to happen. That’s the architecture right now. Crawford’s essays sit with this fact long enough that you feel its weight. She doesn’t race past it to deliver solutions. She lets you understand that this isn’t a technical problem requiring technical fixes. It’s a political economy problem, which means the solutions live in different domains entirely. The writing style does this work. She builds her argument brick by brick, and each sentence assumes you’re paying attention.
The Hunger to Understand Is Real, and Publishers Know It
A Gallup poll from January 2026 found that 61 percent of American adults wanted to understand AI better but didn’t know where to start. That number haunted me when I read it. It’s not a majority expressing polite interest. It’s approaching two-thirds of the country admitting they feel lost in conversations that affect their work, their kids’ education, their privacy. That’s a readership looking for guidance, and it’s hungry.
The Frankfurt Book Fair 2025 trends report documented this appetite across the publishing industry. Over 400 titles in the AI and society category alone. Some will be worthless. Some will genuinely surprise you. Your job as a reader right now is to move slowly through what matters and let mediocre work fall away. The good books reward patience. They’re written by people who have thought deeply about what they’re trying to communicate, and you can feel that care in the sentences.
What Makes a Good AI Book Right Now
I’ve been thinking about what separates the books that actually clarify thinking from the books that just capitalize on anxiety. The best ones are specific. They trace particular decisions to particular consequences. They name names. They explain technical concepts without pretending they’re simple, but without burying you in jargon either. They take your intelligence seriously. They also acknowledge uncertainty. If someone is writing about AI in 2025 with total confidence about the future, they’re probably selling something.
The sentences matter more than usual right now because the topic is genuinely difficult. An author working through these ideas needs to be precise. That precision is what distinguishes a book that deepens your understanding from one that just confirms what you already feared. Read slowly. Underline passages that surprise you. Argue back with the text. This kind of reading takes time, but we’re in a moment where that investment pays off. The conversation about what AI is doing to society is happening, and the books being published now are shaping how we think about it.
If you’ve started reading anything this year that cracked open how you understand this landscape, I genuinely want to know. Text me. I’ll have opinions. More importantly, I’ll want to hear yours.