Nexus, The Revenge of the Tipping Point review: How Harari, Gladwell miss the mark

The book became a runaway success in America and the rest of the world. Arguably, it contributed in creating a new class of bestselling authors — generalists who condensed academic lessons and applied them broadly (some would argue, too broadly) to real-world situations. Thomas Friedman, Jared Diamond et al are good examples of this genre of writers. One of the most popular generalists in the world today is the Israeli medievalist and writer Yuval Noah Harari, the author of the bestselling Sapiens (2011).

In recent months, both Gladwell and Harari have published new books: Harari’s Nexus: A Brief History of Information Networks from the Stone Age to AI and Gladwell’s Revenge of the Tipping Point, a sequel in spirit to his 2000 bestseller. Over the last couple of weeks, I read both titles. Sadly, together they once again prove why the great generalists share a common weakness — over-generalisations, sweeping statements and a tendency to cherry-pick data that retrofits their conclusions. These books are not all bad: they do have occasional moments of insights, not to mention genuinely intriguing data points every now and then. However, on the whole they do bite off more than they can possibly chew.

Good math, bad journalism

In the introduction to Revenge of the Tipping Point, Gladwell says that he wanted to write the book as an ‘upgrade’ to his own methods. Ironically, most of the book feels like it is stuck in the year 2000 itself, when anecdotal extrapolation and marketing-Lite buzzwords ruled the nonfiction charts.

Also read: Pico Iyer’s new book ‘Aflame,’ out in January next year, to explore the power of silence

“In Revenge of the Tipping Point, I want to look at the underside of the possibilities I explored so long ago,” Gladwell writes. “If the world can be moved by just the slightest push, then the person who knows where and when to push has real power. So who are those people? What are their intentions? What techniques are they using? In the world of law enforcement, the word forensic refers to an investigation of the origins and scope of a criminal act: ‘reasons, culprits, and consequences.’ Revenge of the Tipping Point is an attempt to do a forensic investigation of social epidemics.”

The fight to legalize same-sex marriage in America, a spate of suicides at an elite school, Miami’s struggle with institutional corruption — these are some of the disparate topics that Gladwell offers up in his book. The string that ties them all together (or struggles to, anyway) is America’s opioid crisis, led by medicines like Oxycontin that proved to be highly addictive and ruined millions of lives across the country.

The suicide-epidemic story, in particular, shows us how and why Gladwell has become such a frustrating writer. He begins with giving us a sincerely written account of one of the last suicides, and then cleverly starts revealing the common patterns across these days. However, Gladwell forgets that these anecdotes were supposed to be a ‘human interest’ counterpart to all the data around suicidal ideation that he keeps throwing in reader’s faces. It’s not like the data isn’t relevant or doesn’t provide food for thought — the problem is that Gladwell all but forgets the human beings behind these facts and figures. It is a classic case of ‘good math, bad journalism’. It makes it that much more difficult for readers to empathize with these people because we know so little about them other than the circumstances preceding their very last moments. In chronicling their deaths, Gladwell glosses over their lives. In his devotion to counter-intuitiveness, Gladwell ignores his own blind spots.

Also read: Travellers in the Golden Realm review: How Mughal India captured European traders’ attention

As a rule, if a writer is saying “Are you kidding me?” (Gladwell delivers 3-4 variants of the phrase across the book) to describe his incredulity at something, the emotion is probably unearned.

Underwhelming overestimations

Yuval Noah Harari’s latest, Nexus, is a study of information networks from medieval times up until the most contemporary iteration of the idea, which is Artificial Intelligence (AI). His mandate here is to “cure” what he reads as mankind’s “naïve view of information”. But even if one were to take Harari at face value, it would have to be under the assumption that most people function daily without having sophisticated ideas about information and its true nature.

Harari’s logical fallacies are a bit more sophisticated than Gladwell’s. His mode-of-operations is to construct elaborate, easily demolished strawmen that can then be used to bolster his core arguments and smooth over any contradictions that might pop up. Here, for example, he says that the free and fair flow of information is almost always a net positive (a self-evident truth). However, he goes on to make a startling claim in the same paragraph — namely, that a large number of people around the world believe that technical or scientific knowledge will also somehow lead to political emancipation.

I cannot emphasize enough how stupid this notion is. There are entire academic departments around the world geared around the idea of de-coupling technical knowledge and political agency. “Knowing that e=mc2 usually does not resolve political disagreements”, Harari writes. This is a statement that is as true as it is useless, mostly because no serious academic can possibly believe the opposite. And yet Harari labours under the delusion that we the readers are the truly lost ones.

Harari is similarly misguided on the issue of AI. Showing the kind of tech-alarmism that used to be in vogue 20 years ago, he basically draws a straight line between the kind of Large Language Models (LLM) widely used today — and the totalitarian, sci-fi tech seen in mainstream Hollywood films like Terminator or Transformers.

Also read: 2024: The Election that Surprised India review: When Modi magic wane

The problem is, ChatGPT or any other LLM-based application so far hasn’t even shown a fraction of the power that Harari allots them in Nexus. These large-scale plagiarism machines can barely summarize a journal paper without churning out frequent ‘hallucinations’ — false and/or made-up data, citations that do not exist, authors conjured out of thin air et cetera. But to Harari’s eyes, this level of progress (or lack thereof) is enough to sound the alarm bells. Regurgitating high-school level ‘servant becomes the master’-style arguments about this technology, Harari actually does a grave disservice to those of us who criticize LLMs for the right reasons (they are stealing the work of artists and knowledge creators while simultaneously relegating them to the margins of industry and capital-networks).

Harari writes, “Knives and bombs do not themselves decide whom to kill. They are dumb tools, lacking the intelligence necessary to process information and make independent decisions. In contrast, AI can process information by itself, and thereby replace humans in decision making. AI isn’t a tool — it’s an agent.”

This passage is typical of the book, to the extent that it ends up committing the same logical fallacies it warns us against. By saying ‘AI isn’t a tool, it’s an agent,’ Harari ironically ends up shielding the tech CEOs, venture capitalists and investment bankers who are funding the degradation of knowledge and culture via LLMs. It’s like saying Sam Altman does not have the ‘agency’ when it comes to ChatGPT’s failures, it’s the fault of the software itself.

Both Harari and Gladwell have positioned themselves in the media and in popular discourse as fearless truth-tellers, as narrators of ‘difficult’ or ‘suppressed’ stories. But as their latest books reveal, this positioning itself is the result of logical fallacies, over-estimations and sweeping statements that often fail to correspond to the on-ground facts. Which is a shame because at the premise level, both books had a lot of potential. One hopes that these two writers use their considerable gifts with a little more discretion the next time around.