Alien Intelligence

Yuval Noah Harari is an important voice in modern discourse. As a historian, he brings a perspective that relates how patterns of human behavior in the past are relevant to the present. His latest opus, Nexus, is an insightful and sometimes frightening look at what the near future could hold.

Prophet, said I

Two things I appreciate about his writing. One, he cautions us against magical thinking, for instance, that the arc of history tends toward justice. Two, he eschews the idea that “history repeats itself” as the old adage says. As a historian, Harari does not wish to take up the mantle of a prophet, rather that of a guide to help us think out the possibilities that could arise from the information networks we are currently building, like social media, and especially AI.

The title, Nexus, references his main thesis, that the ability of Sapiens to achieve dominance on the planet is tied to our language ability–that one mind can create reality in another mind by means of communication. Through language we form information networks that bind large groups of people to a set of shared goals. Leaders are those who gain control over the information in the network. Through the mind-to-mind “nexus” people are able to forge new shared fictions that people agree to regard as real. This builds upon a key thesis of one of his previous works, Sapiens, referenced above.

The trouble is the goals are evil as often as they are good, and he spends great deal of time outlining how over the course of human history that people have used information networks to create what he calls “intersubjective realities”–ideas that do not have grounding in empirical truth, but nevertheless are real inasmuch as they affect the actions of fellow humans.

I very much enjoyed how he takes us through the history of information networks, using the Sumerian invention of cuneiform as an early example. The mere ability to store information leads to an information retrieval problem as the network of tablets grows. Bureaucracies form, and assume a great deal of control over people’s everyday lives. The ability or inability for individuals to store and retrieve relevant documents can mean the difference between freedom and prison, even life or death.

Malleus Maleficarum

As an example of how information can be used nefariously, leading to death, Harari discusses the emergence of the printing press in late medieval Europe.

Printed works such as the Malleus Maleficarum, i.e. “The Hammer of Witches” and other derivative and related titles, formed a sinister network of misinformation, motivating generations of Europeans to believe in and scapegoat witches. Neighbors would defame marginalized people. Their witchcraft was an affront to God, and God has thus visited upon us all manner of problems like plagues, bad harvests, and the untimely deaths of loved ones. Witches must be punished and eliminated if the problems are to stop.

Today, we imagine ourselves as more evolved and click our tongues at the foolishness of our ancestors for believing in witches. We shake our heads, exasperated at the immense injustice of it. Most of us regard the notion of prosecuting individuals, or even whole families, for the crime of “witchcraft” as risible.

Nonetheless, in the early modern period, an information network facilitated by the invention of printed books caused it to loom large in the shared “intersubjective” reality, in the minds of other sapiens not too different from us, at an immense cost of human suffering.

Alien Intelligence

As someone who makes use of AI on a daily basis, I find it to be a real accelerator to productivity. It’s like having Encyclopedia Brown in your pocket, able to answer questions that I can casually compose on any subject, quickly. Not only that, but it is capable of understanding context of my question without me having to make it explicit. This saves tremendous effort and time trying to figure out what relevant details to provide to frame the question.

Chat GPT, Microsoft Copilot, Claude, these all provide relevant answers very quickly. Using these tools is much more efficient than paging through Stack Overflow posts–no carping by persnickety gatekeepers, no spam, no ads, no unending list of links to visit to see if they might pertain to what I need to know.

Harari makes a point to refer to AI, not as “artificial intelligence” but as “alien intelligence.” Most of the second half of the book focuses on this notion, that the intelligence we are creating is not so much artificial, but more importantly, it is alien to us. Computers process information in a means that is impossible for humans. They are able to consume vast amounts of it, quickly, and to never forget or miss a detail. This is in stark contrast to human intelligence, which relies on intuition and emotions rather than vast quantities of data to make decisions.

There are several chilling descriptions of how this technology is already being put to use in ways that could eliminate the notion of privacy, and through that, the notion of individual freedom. In particular, he notes how the Iranian government uses computer vision to monitor women who might be violating the country’s dress code, even automating enforcement.

Here in the US, as our own government onboards the use of AI for more and more tasks, it is chilling to imagine what could lie ahead of us. I wonder if it is even possible to avoid some very dystopian outcomes.