After NEXUS: Three Books That Reshaped How I Read AI

Artificial intelligence is often discussed as if it were one unified force, rather than a complex mix of ideas, interests and philosophies. The author had been reading technical papers and vendor whitepapers, believing that a deeper grasp of neural networks or transformer architectures would clarify the implications of AI. However, reading Blake Leno's NEXUS book changed their perspective dramatically.

NEXUS traces how connectivity itself rewires human cognition, from writing to the internet, and reveals that AI is not simply a technical issue, but a social one. The algorithms matter far less than the power structures they embed. Once you understand this, three other books become crucial: Yuval Noah Harari's Sapiens, Cathy O'Neil's Weapons of Math Destruction, and Bruno Latour's We Have Never Been Modern.

Harari's Sapiens shows that large-scale cooperation relies on shared myths, such as money, nations, and humans being inherently rational. Recognizing this, we can see how myths shape AI systems at scale when algorithms make decisions about credit, hiring, and criminal justice. O'Neil's Weapons of Math Destruction exposes how algorithmic systems, even well-intentioned ones, amplify inequality and lock people out of opportunities.

The power asymmetry is structural: algorithms are deployed without understanding, consent, or accountability for failure. Latour's We Have Never Been Modern argues that our intellectual tradition rests on a false divide between nature and culture, science and politics, fact and value. We cannot separate them, as every scientific fact is entangled with social interests, and every technology embeds politics.

When we build AI systems, we are not separating intelligence from values - we are weaving them together, then claiming neutrality. This insight highlights that we must acknowledge every system we build is a hybrid of technical and social elements, and we must govern it as such. These three books converge on the idea that AI's crisis is not about intelligence, but power, myth, and the stories we tell ourselves about objectivity.

The author notes that these books ask the right questions, not for answers, but for a deeper understanding of the permanent tension between what systems can do and what we collectively decide they should do. This tension requires ongoing negotiation and admitting that every choice about AI's development and deployment is about shaping the kind of society we want to live in.

The author emphasizes that technical literature on AI is insufficient alone; we need engineers who understand the mathematics deeply and can ask hard questions about systems, as well as people who understand the social consequences well enough to demand accountability. Engineers must read Latour, policymakers must understand neural networks, and everyone needs to stop pretending AI is a neutral tool waiting to be aligned by the right incentive structure.