The AAA-ICDR Institute, the new think tank at the AAA, will also coordinate AAAiLab™, where, since last year, the Association has been collecting and sharing insights on AI in ADR and on how we are integrating AI into our operations.
For all the technical challenges AI presents, what we are building also entails complex philosophical choices that will be just as important to get right. How do we leverage this new mode of intelligence without extinguishing the spark that ignites trust, and empathy, and all the intangibles that make us human? Resolution of conflict—disagreements, lawsuits, community strife—is a highly personal matter for all involved. And preserving space for human intelligence to guide its new “artificial” counterpart will be important in carrying forward our responsibilities as an ADR administrator.
Perhaps AI is not so different from other major advances in technology that humans have confronted over thousands of years. And that would be a good thing, meaning that we are not completely adrift, that we can reach into the past and learn from how others engaged with discoveries that gave them, like us today, the potential to exceed natural human ability in shocking and disorienting ways.
An example: Western art from roughly 1500 until 1700. Earlier European painting tended to look flat, with almost cartoonish depictions of reality:
Sano di Pietro, Saint Anthony Distributing His Wealth to the Poor, c. 1430/1435 (National Gallery of Art).
Giovanni Bellini, Portrait of a Young Man in Red, c. 1480 (National Gallery of Art).
Then, over the course of not much more than a century, there was a blossoming of what must have seemed at the time stunningly realistic paintings of people, places, and things. Faces brimming with emotion. Muscles straining in three dimensions. Fabrics draping in complex ways, shimmering in the light, and retreating into shadows. Household decor, mirrored surfaces, and sun-filled rooms that look much more like what we see with our own eyes:
Antonis Mor, Self-Portrait, 1558 (Uffizi Gallery).
The step change in Western painting that coincided with the High Renaissance might be attributed to the rise of humanism, better understanding of anatomy and perspective, and unimaginable wealth driving insatiable demand for beautiful objects, among other factors.
However, one theory, known as the Hockney–Falco Thesis,1 posits a simpler and more direct explanation. Maybe many of these artists leveraged the high technology of that time to create artwork that was remarkably more realistic than anything that had come before. The theory goes that painters began using optics—mirrors and glass lenses—to cast real-life scenes onto their canvases, and then essentially traced those projections in paint.
Johannes Vermeer, The Milkmaid, c. 1660 (Rijksmuseum).
The evidence supporting this theory is compelling. And it leads to a conclusion that is less obvious than we might expect when confronted with a tool that enhances our abilities to an extraordinary degree: Humans can use a technology well or poorly. For every Vermeer that looks both photographic and painterly—the thought and skill and effort that went into his work is undeniable—there are photo-realistic but clumsily executed works by other artists from the same period. The perspective seems disjointed, the image is reversed so everyone appears to be left-handed, or the artist’s technique simply feels soulless and mechanical. Having access to this technology would not have been enough. The most striking art of the period was not merely some “paint-by-numbers” exercise.
And then there’s an artist like El Greco. He lived relatively late in this period, but it seems hard to argue that he relied on optics and projections, given the unnaturally elongated human figures characteristic of his work. Even so, his art comes across as far more advanced, in its own unique way, than most works painted before 1500. Again: We can use a certain technology well or poorly—or avoid it entirely when it does not serve our purpose. Maximizing its use at the exclusion of all else is never the only option available to us.
El Greco, Saint Jerome as Scholar, c. 1610 (Metropolitan Museum of Art).
Bringing this back to AI: The human hand—and not this mind-boggling new technology—is still the true difference-maker. Our taste. Our values. Our skill. The care we put into the things we create with it. So, when we are using generative AI to help draft an article or email, or to build a chatbot, or to automate a complex process, it might do to remember that the result need not exist wholly apart from us. Rather, we can and should incorporate much of ourselves, much that is human, into the work. And we can communicate, as transparently and explicitly as possible, why and how we struck the human–AI balance that got us there.2
ADR processes will continue to impact real lives here in the real world, no matter how advanced and pervasive AI tools become. Everyone who engages with what the AAA builds, and what we say about what we build, may be able to sense whether we used AI well, poorly, or not at all—and they may care a great deal one way or the other. “How will people feel about this?” That is the starting point. Is it ethical? Is it trustworthy? Can we explain how it works? Is it improving how people resolve their disputes? Our essential humanity can and should serve as the foundation for AI-driven services that remain familiar and authentic to those who interact with and rely on them.
The theory bears the names of David Hockney, one of the world’s most influential living artists who wrote a book on the topic titled Secret Knowledge (also the title of a BBC documentary on the subject), and Charles M. Falco, a physicist and expert in optics.
I almost “avoided AI entirely” in writing this, only asking GPT-4o to flag any overused words or phrases. That balance seemed appropriate here, given the topic, as did the most-used words: human(s) (9), AI (8), and technology (6). My prompt: “In the article below, which words (other than common words) and two- or three-word phrases have I used more than once? Give me a list with the frequency of each.”
 | A guest post by
|