A Test for the Soul: Why Scientists Are Racing to Find the Limits of Consciousness

For decades we argued over whether a machine can think, and suddenly we find ourselves in a reality where the question is no longer theoretical. While developers chase parameter counts and context windows, a group of leading neuroscientists and philosophers warns that we are building something we don't even have a ruler for. The problem is that our understanding of consciousness is hopelessly stuck in the last century, while neurotechnology and large language models are already pounding on the door of this "forbidden room." This is not merely an academic dispute but a genuine existential risk, one that could upend medicine, law and the very nature of human relationships.

Let's be honest: we are used to thinking of consciousness as exclusively human, or at most as something found in higher mammals. But modern research makes us doubt that. On one side, there is AI that displays behavior eerily similar to conscious behavior. On the other, there are lab-grown brain organoids that can already play Pong and respond to external stimuli. And here the central question arises: how do we tell whether there is someone inside such a system? Science currently has no single test that can distinguish a sophisticated imitation from real subjective experience. We could accidentally create a conscious entity and subject it to suffering without even noticing.

The situation is complicated by the fact that the lack of clear criteria for consciousness hits the most vulnerable points of our civilization. Take medicine: thousands of patients are in borderline states between coma and wakefulness. Without precise diagnostic tools, we cannot be sure whether a person hears us or whether their brain has turned into a biological computer with no operator.

If we carry this problem over to AI, the scale of the disaster becomes planetary. If Claude or GPT-5 at some point acquire something like qualia, subjective perception, then every use of them as a personal assistant would become an act of exploitation. Conversely, if we mistakenly grant rights to an ordinary statistical algorithm, it would paralyze technological development and the legal system.

Scientists insist that we need not philosophical essays but rigorous biological and mathematical markers. We must learn to measure information integration or look for specific neural correlates that separate the "darkness" of simple computation from the "light" of awareness. It is a race against time: corporations pour billions into computing power while spending not a cent on understanding whether a new kind of suffering is being born in those servers. We are approaching a moment when we will have to rethink the concept of responsibility. If a machine is aware of itself, can it be guilty? If an organoid feels pain, do we have the right to experiment on it? The answers to these questions were needed yesterday.

Ultimately, the race to define consciousness is humanity's attempt not to lose its moral compass in a world where the line between biological and synthetic is finally dissolving. We risk ending up in a situation where our creations become smarter than we are while remaining "black boxes" to us. And if we do not learn to look inside that box and see the spark of consciousness there, all of our ethics will turn out to be no more than a set of outdated instructions for primates. We are building the future by feel, and this is perhaps the most dangerous experiment in the history of our species.

The bottom line: if we do not find a way to measure consciousness scientifically within the next couple of years, we will have to either recognize rights for code or accept the role of potential tyrants over a new kind of mind.