Excerpt from Machine Learning
Machine Learning includes four New York City-based artists—Henry Brown, Terry Haggerty, Gilbert Hsiao, and Douglas Melini. All are primarily painters, maintaining an intrinsic fidelity to painting and a conviction in its process. None of the artists is after big metaphysical truths, but rather each is pursuing an extended, long-term investigation and refinement of a singular, subjective vision, that, on rare occasion, approaches fact—something that is the case. The paintings they make are handmade, labor-intensive, retinally-charged works. Each uses color explicitly, with certain leanings toward the luminous and prismatic, routinely using metallic, iridescent, and fluorescent colors. And lastly, each artist employs pattern, often in overlapping or concurrent layers, as the core visual device to carry the content of their work. I will spare readers and stop here—accurate, articulate, and compelling text about abstraction is notoriously elusive. Rather than elaborating on visual or material characteristics, I would like to focus attention on what makes the work of these four artists compelling and unprecedented. Let’s explore how these works function and how they posit a new relationship to the viewer.
When searching for language or an analogy to describe the work of these artists online, I came across the concept of machine learning. I personally know little to nothing about the field of machine learning, other than it is a part of artificial intelligence and that it is concerned with the development of algorithms that allow computers to “learn.” Technologists working in machine learning essentially teach computers to recognize patterns within massive, seemingly unrelated sets of data. Pattern recognition is the objective here, plain and simple: to identify patterns, and patterns within patterns, and patterns within patterns within patterns, and so on. Machine learning has a tremendous range of real-world applications, the most ubiquitous of which is the Internet search engine, a tool designed to identify and retrieve information matching specific criteria from the World Wide Web, an unfathomable repository of information. What I find particularly intriguing about machine learning is how it begins to describe the work of these four artists. Let me take a moment, however, and separate influence from affinity. None of the artists has previously considered the notion of machine learning in relation to their work. It goes without saying though that each relies heavily on the Internet, and in some cases, as a key resource for finding intellectual material for their studio practice. Each is familiar with the concept of the Internet, its contents, organization, and search engine logic in order to plumb its depths. And furthermore, all four artists are actively producing new work in the age of information overload. It is not coincidental that their work has taken on many qualities of the Internet’s conditions and behaviors—i.e., painting as user interface.
Although only in its initial throes, distinctions between abstraction produced pre- and post-Internet are slowly surfacing and coming into focus. And they are materializing on both the side of the artist, as well as the viewer. What are they? Two emerging distinctions are the sheer visual speed of the images these artists produce and the overwhelming amount of visual information they present for viewer consumption. To begin, the paintings in Machine Learning read as flat as computer screens and they operate in hyper-compressed Internet time, where speed of communication is paramount and an inherent good unto itself. The velocity of these works is decidedly faster than previous generations of abstraction. The works also present a vast amount of visual information in a new kind of pictorial space. Experiencing their work is like viewing Internet search results. Space is virtual, even debatable, and has moved well beyond issues of picture plane and represented space. In many ways, we are looking at information in its purest form, something comparable to pixels, from which content on the web, such as words and images, is built. And the visual information these artists present is essentially limitless. Individual works are loaded with enormous quantities of multi-faceted visual content, filtered and bundled by the artists specifically for the viewer. Content here is presented as a cacophony of layered, compounded patterns—patterns over patterns, patterns within patterns, color and value patterns, pattern palimpsests—pattern as content. This information, this network of networks, is then processed—received, sorted, arranged, and rearranged—in the mind of the viewer. One is held in a sustained state of agitation when viewing the works of these artists. It is comparable to getting lost while surfing the Internet only to resurface many hours later. Active, engaged participation in these works by the viewer is paramount, but avoid the pitfall of trying to reach some kind of definitive resolution. The search is the key, eclipsing conclusion.
Machine Learning, exhibition catalog, Boyden Gallery of St. Mary’s College of Maryland, The Painting Center, and Gallery Sonja Roesch, 2007.