A Human Algorithm: How Artificial Intelligence Is Redefining Who We Are

Image of A Human Algorithm: How Artificial Intelligence Is Redefining Who We Are
Release Date: 
October 1, 2019
Reviewed by: 

The urgent task is to rapidly reclaim and amplify the best of ourselves, and this book is a step on that journey.”

One may want to live in the holistic, evolved world Flynn Coleman envisions in A Human Algorithm and sincerely hope that we can somehow find our way there collaboratively and in a spirit of dignified grace. It’s important to say this up front, firstly to breathe some additional life into the vision that might help it manifest, and secondly to indicate that any of the argument’s shortcomings (most notably concerning exactly how to achieve the vision) don’t undercut this. 

At the same time, it isn’t always clear that those who currently exercise undue power and influence in our world are similarly inclined. Cynically speaking, they tend to centralize their dominion, insulate themselves from challenges, and move ahead into a brave new world without procuring the consent of the governed (so to speak). Likewise, many of the ostensible advantages their work provides—whether via productivity tools, personal assistants, or infotainment apps—often seem more about capturing our innermost energies and homogenizing our experiences for abjectly pecuniary purposes.

Coleman clearly grasps the gravity of the moment and the complex forces at play, describing in eloquent terms not only the changes already at hand but the ones right around the corner in the “next wave” that will push the frontiers of human control and consciousness. Yet despite the technological “quantum leap” in the offing, Coleman cogently observes in her introduction that “many of us remain head down, staring at a brightly lit screen as the tectonic plates of history and technology are shifting beneath our feet.” Unprecedented change is plainly at hand while we stare into our palms.

Left unanswered in this lamentation, however, is how we are to retain our humanity and bring a “more inclusive set of stakeholders” into the discussion of how to infuse artificial intelligence (AI) with humane values (which is the primary goal of the book), when on the other side of those brightly lit screens are sophisticated teams of designers intent on keeping our heads down and their profits up. Coleman talks about society’s “technological fears” and the fact that we’re “alarmingly unready”—but perhaps the bigger problem is that we’re not actually afraid enough.

Against this realization, A Human Algorithm is intentionally positioned as an optimistic book, one that recognizes the challenges but affirms our human capacity to rise up to them. History shows (and the book details) that people have a marked capacity to integrate new technologies into their lives, even as both horrific and honorable applications may ensue. The Digital Age (which Coleman wisely pronounces as “now ending”) may be unique for the rate and magnitude of the changes being experienced and the unprecedented nature of the frontiers being pushed, but the general patterns remain intact.

The problem, of course, is that if we build machines in our image—or at least based on our values and behaviors—all that makes us human will be “mirrored in the AI.” This potentially means having machines that can just as easily kill us as save us, heal or make ill, uplift or denigrate—just as we see in the human-on-human realm. Coleman cites many examples of these dangers, but virtuously insists on accentuating our capacities for empathy, ethics, and equity as well. In this view, the task is to infuse AI with as much of the good as is humanly (and mechanically) possible.

This otherwise unobjectionable notion brings up a thought that has probable sleep-impinging implications. You know all that data being collected on all of us, all the time, by just about every automated device and artificial intelligence in our midst? The information compiled for a single individual might fill a small aircraft hangar if printed out, and the sum total of human behavioral metrics might stack up into outer space. What if machines acquired that massive, unvarnished data as the baseline for assessing and/or emulating humankind? You see the point.

Then again, this is precisely Coleman’s point, inverted: the titular “human algorithm” is the sum total of who we are, all of us, “the DNA of our humanity” and “the philosophical center of ourselves,” as Coleman writes in the conclusion. This includes our foibles and failures as well as our capabilities and kindnesses. And perhaps having this complexity mirrored back to us will indeed become “the chrysalis for the biggest metamorphosis humankind will ever experience,” leading us ultimately to “reimagine our place in the universe [and] our connection to all things.”

One might truly want this to be the case and believe that saying it out loud—backed by the extraordinary analytical pastiche comprising the bulk of A Human Algorithm—is an important affirmation and a necessary step toward any possibility of making it so. This moment in time may well be our “best and last opportunity” to reconnect with ourselves and the world around us, and to enshrine the best of our values into the coding of the machines that will outstrip our abilities in short order. We have to get this right if anything of us will be left.

By closing on an expressly hopeful note—after tracking some of the most pernicious prospects along the way—A Human Algorithm represents a choice: descend into darkness, or “bend collectively toward the light.” Indeed, this has always been the inflection point before us, individually and as a whole. Now, with machines ready to assimilate and instantiate the data we’ve compiled on ourselves, we’re on the cusp of seeing our choices writ large, immutably and eternally. Setting the baseline coding for the future is the sacred responsibility of all of us here and now.

In A Human Algorithm, Flynn Coleman admirably has uploaded a clear data point on the side of the good. The book is not, however, an idealistic portico onto a rose-colored tomorrow; Coleman is keenly aware of the existential implications and the narrow window of opportunity at hand. The urgent task is to rapidly reclaim and amplify the best of ourselves, and this book is a step on that journey. In addition to urging people to read it—especially those directly involved with writing the algorithms that will set tomorrow’s template—one sincerely hopes the machines will do so as well.