The Point of Intelligence
There is still a lot of disagreement over how, exactly, photography was received by artists when it first arose as an invention. Along with the general public, many artists took notice when the first daguerreotype appeared in the late 1830s. But they disagreed about what bearing, if any, the technology ought to have on art. I want to consider for a moment the version of this argument that says photography created an existential crisis for art and artists: that when photography emerged, many artists understood their work in primarily representational terms. Furthermore, these same artists saw in photography a supreme representational accomplishment, a challenge to the worth of their work that was all the more grave because it could be achieved with minimal skill by the “artist” (photographer). Then, so the argument goes, art started to move down the road toward modernism, which was essentially a set of post-representational innovations that distinguished the purpose of avant-garde art from photography.
I wonder how an analogous story might play out again with writing, knowledge work, and the recognition of chatbots. Our own moment leads me to reflect back on the situation with art and photography almost 200 years ago now, and makes me think that maybe it wasn’t so much the artists who perceived a threat to their work, as it was the public that (re)interpreted art in terms of photography. If large numbers of people see the artist’s work as essentially about representation, about reproduction of reality or things “as they really are”–then in some sense it doesn’t matter what the artist thinks he or she is doing. You can decide that large numbers of people misunderstand your work and continue on doing the same things, but you can also lose your audience in the process. Even if the reasons for the change are hard to discern, it seems that art underwent a paradigm shift from within a world that could be photographed.
In the same way that artists pursued a multitude of ends at the time photography arrived, there will never be any kind of agreement on what writing is or is “doing.” Still, regardless of what writers think they are doing, automated methods will find a way to produce a refracted copy of it–at least some of it. But automation like a chatbot has a very different way of presenting what it does to the public. For example, automated writing “responds” to a “prompt,” it “completes tasks,” or “answers questions.” In the same way that much of the art world was collapsed into the self-presentation of photography, writing risks losing some of the rich account of itself when it is presented with an apparent copy by machine.1
Chatbots push us a little further into a model where writers have goals, where they have “information they want to communicate” (where is the information if not in the writing?). I wonder if we might see something like a photographic reckoning for the writing world today, where the apparent similarity with automated methods leads to a profusion of new genres and self-justifications for, say, literary writing. And could an analogous re-evaluation occur in more utilitarian writing forms (e.g., the professional memo, advertising) as well? Might all forms of signifcant writing need to situate themselves on new conceptual footing, to account for the investment of human time and energy in the shadow of the machine?
- Art, too, faces another version of this with chatbot-likes tools: if suggestive new artwork can be generated with a simple concept typed out in a prompt, does this further threaten the representational justification for art? ↩