When a German philosopher traced many aspects of human thought back to a profoundly earthly view of mankind, defining it as “human, all too human,” he carved a sharp and nihilistic break with the Romantic tradition that had long flirted with the eternal, the transcendent, the divine. But when it comes to art, how decisive is the impact of the imperfection of the human hand?
A large part of classical art was built around an idea of perfection, often overlapping with the divine dimension: unsurprisingly, religious subjects dominated for centuries. It took a very long time before anyone began to perceive something genuinely compelling in error, deviation, and imperfection. The twentieth century was likely the apex of this shift in awareness: a continuous sequence of ruptures, dissonances, and revolutions that cut across all art forms, from the visual arts to music, transforming divergence into value.
Today, however, living within a civilization permeated by artificial intelligence forces us to confront new paradigms. The benchmarks that govern our work and the creative models populating the infosphere are increasingly synthetic, polished, seemingly perfect. This is precisely what we demand from machines: not to make mistakes. But what would happen if machines began to err? Would we truly be capable of finding an algorithm’s mistake fascinating, as we do with a crooked Nirvana guitar solo? To what extent are we willing to tolerate imperfection when it is no longer human?
Let us push the argument to its extreme. It has been demonstrated that self-driving cars cause far fewer accidents than those driven by humans. And yet, how prepared are we to accept that an algorithm might “decide” to run over a pedestrian, perhaps based on a utilitarian calculation that saves three others? Similar incidents occur every day: fatigue, alcohol, distraction. There are trials, blame, responsibility. But if such a decision were made by a machine, how deeply would it destabilize the entire system of trust on which the widespread adoption of AI in everyday life depends?
The same applies to medicine. It is now well established that, in certain diagnostic fields, machine learning algorithms can identify tumors and pathologies that escape the human eye. And yet, even here, the question remains open: how willing would we be to accept that, even once, a wrong assessment could endanger a person’s life?
Perhaps this is because imperfection is, at its core, the defining trait of our species. A primordial gift, if we wish, from a mischievous deity. Human beings are born in pain and blood, and live by frantically chasing brief moments of balance, inevitably followed by new fractures. This is likely what we also seek in artistic representation: not an artificial perfection, but something that resembles us, betrays us, exposes us.
If one day we were to eradicate every disease, destroy every microbe, extend human life to hundreds of years, we would likely find ourselves living in an absolute, lethal boredom. And, in all probability, no one truly desires such a world. As Bukowski wrote: “We are here to laugh at the odds and live our lives so well that Death will tremble to take us.”
