A Sartrean Criterion for AI Consciousness

​In this article, I will attempt a small intervention in contemporary artificial intelligence discussions with an old philosophical example. Our hypothesis is this:

​”Artificial intelligence is very powerful at defining and reproducing existing patterns. However, experiencing that which is not there, absence, negation, and nothingness, as a lived condition belongs exclusively to human consciousness.”

Jean-Paul Sartre, in his famous “Cafe Example” from his 1943 work “Being and Nothingness”, describes the experience of someone entering a crowded cafe to find a friend Pierre. The space is filled with other people. But Pierre isn’t there. Eventually, for the seeker, the “non-being” of Pierre becomes a more dominant experience than all the “present” things in the room. Nothingness is not an object in the external world; it is an experience projected onto reality by human consciousness. According to Sartre, this can sometimes push all that exists into the background.

Sartre’s perspective suggests an interesting criterion for whether artificial intelligence can possess consciousness in the sense we understand it. ​Artificial intelligence is saturated with data; in fact, it is a hub for reproducing existing patterns. It processes data, patterns, definitions, labels, and existing entities. It operates entirely within the realm of “what is there.” It can successfully identify what an object is, or even state that a specific object is not present in an image. Yet, this state of “absence” remains a result drawn strictly from available data. Machine learning may utilize negative weights or inhibitory connections, but these are all parts of the model’s own mathematical representation. What I take from Sartre, however, is a distinction beyond this. For a human, nothingness is not merely a negative value; it is a unique experience constructed by consciousness.

​For instance, missing someone is precisely this kind of phenomenon. In Sartre’s example, I am trying to unpack the difference between the meaning of Pierre’s absence in a cafe and the lack of pixels belonging to Pierre. ​When you feed a photo of a cafe into an AI model, it identifies the tables, chairs, and people. If Pierre is not in the photo, the model simply returns Pierre: False. The non-presence of Pierre has no distinct ontological meaning for the model compared to any other object. ​Yet, when human consciousness walks into that same cafe, Pierre’s absence is not merely a “0”. That non-being can transform into an experience far more forceful and present than all the physical entities in the space. For humans, absence can sometimes become more visible than presence.

Do human memories not also rely on signals that once existed? Sometimes yes, but sometimes, it feels like no. I remain undecided on this point. Even so, in the framework proposed by Sartre, nothingness and non-being are not derived from a dataset. They form a distinct layer of meaning brought into the world by human consciousness.

​We are accustomed to defining everything through data, positives, and existing entities. So, grasping this concept of “nothingness” can be challenging. For example, when you are alone, you might feel bored; you want to do something, yet you abandon every option that comes to mind. Your friends might tell you, “You are thinking too negatively.” According to Sartre, it is precisely through this act of negation that human beings can gain freedom.

Recognizing the pure void left by something that is not there and making choices within that void is one of the most defining characteristics of human consciousness and freedom. In addition, it is surely a human condition to miss something. Maybe we should remember this more frequently.

​In summary, artificial intelligence establishes mathematical relationships among patterns within a dataset and defines what is absent strictly through these relationships. Humans, on the other hand, can live through the emptiness left by non-being. They can bring the absence of something that never existed or something that has departed into the world as an experience more commanding than existence itself. Perhaps Sartre’s concept of “nothingness” can help us propose a different benchmark for consciousness in AI debates. At the very least, we may now consider boredom not necessarily a bad thing.