Perplexity: An Untidy History of AI Across Four Books By Trevor Quirk Published in The Hedgehog Review, Summer 2025, Issue "Lessons of Babel" – Book Reviews --- Overview The article explores the complex and often chaotic history of artificial intelligence (AI), emphasizing its technological evolution, cultural hype, and the mixed narratives presented by leading AI commentators and prophetic authors. It reviews four prominent books on AI: Nexus: A Brief History of Information Networks From the Stone Age to AI by Yuval Noah Harari The Singularity Is Nearer: When We Merge with AI by Ray Kurzweil Genesis: Artificial Intelligence, Hope, and the Human Spirit by Henry A. Kissinger, Craig Mundie, and Eric Schmidt AI Snake Oil: What Artificial Intelligence Can Do, What It Can Can’t, and How to Tell the Difference by Arvind Narayanan and Sayash Kapoor --- History and Development of AI AI traces its roots to ancient technologies, such as the abacus, automating formal and cognitive tasks. Postwar AI research began with symbolic AI focusing on logic and semantics but stagnated by the early 2000s. Machine learning, especially neural networks inspired by brain structures, gained momentum in the 2010s, accelerated by abundant data from the Internet and GPU computation. Milestones like the 2011 ImageNet breakthrough shifted resources to neural networks, shaping modern AI applications like search engines and social media. OpenAI’s launch in 2015 and the surprise popular success of ChatGPT in 2022 popularized generative AI, fueling a massive commercial boom and hype. --- AI Hype Versus Reality Narayanan and Kapoor’s AI Snake Oil warns of excessive hype surrounding AI, urging public skepticism and distinction between generative AI (content producing) and predictive AI (forecasting real-world events), the latter often failing or unachievable. The authors express concern about the societal and ethical consequences of AI’s popularity, misuse, and misleading commercial hype. The term "AI" has become overused and misapplied, sometimes labeling basic appliances or non-AI systems, which confuses public understanding. --- Critical Examination of Leading AI Commentators Yuval Noah Harari Harari’s Nexus attempts to contextualize AI historically but frequently errs in technical understanding, mixing metaphors and misunderstandings about machine learning and AI’s limitations. His treatment of AI concepts like the alignment problem (how AI goals might diverge dangerously from human goals) is critiqued as misapplied or confused. While critical of "opaque" AI tools like COMPAS (a predictive recidivism scoring system), he downplays their inaccuracy. Kissinger, Mundie, and Schmidt (Genesis) Their book is characterized by grandiose hyperbole, political and philosophical speculations bordering on eschatology, and tech oligarchs’ worldview. They dwell on fatalism and social passivity in face of AI’s rise, arguably legitimizing these attitudes as inevitable or desirable. They emphasize existential risks of AI by comparing it to nuclear weapons, framing AI development as an international arms race. Their discussions sometimes veer into quasi-religious or mystifying territory, invoking divine right and AI as godlike forces. Ray Kurzweil Noted for his techno-optimistic futurism, Kurzweil’s The Singularity Is Nearer builds on the idea of exponential technological progress leading to a "Singularity" by around 2045. His predictions include life extension via AI-controlled nanotechnology, digitization of minds, and AI-human mergers. His work is both criticized for speculative leaps and noted for reflecting a personal quest to transcend death (he is cryonically preserved). Kurzweil uses a mix of scientific data, philosophy, and large numerical claims to foster belief in a near-future AI utopia. --- Social and