As artificial intelligence systems develop, an old question has resurfaced: Can machines possess consciousness? The debate is no longer confined to philosophy or science fiction; it has become part of a broader scientific and ethical discussion that also encompasses animals, embryos, and lab-grown brain organoids.

But a new analytical study published in the journal Neuron doesn't attempt to directly answer whether AI is conscious. Instead, it poses a more fundamental question: Does current science possess the tools to accurately measure consciousness itself? The study was conducted by a team led by Hakuan Lau, director of the Neuroimaging Research Center at the Institute for Basic Sciences, in collaboration with researchers from the University of Montreal and New York University.

The problem of measurement:

Researchers believe that much of the current consciousness research may not clearly distinguish between subjective experience and information processing. This distinction is important because a system, whether a human brain or a computer model, may be able to receive, analyze, and respond to information without necessarily having an internal conscious experience.

As reported in the report, Lau stated that many current theories about consciousness appear to be supported by scientific experiments, but those results may reflect “general information processing” more than consciousness itself. Therefore, it remains difficult to definitively say that these theories actually explain consciousness.

This point complicates the debate about AI consciousness. If scientific tools are not yet able to isolate consciousness from other cognitive processes in the brain, it becomes difficult to use them confidently to judge non-human entities, such as AI systems or non-verbal organisms.

Insufficient experiments

: Researchers criticize some common experimental paradigms in neuroscience, such as visual cloaking, interocular competition, and perceptual threshold tests. These experiments are used to study when something becomes visible or consciously perceived, and when it remains outside of direct awareness.

But the problem, according to the study, is that these tests don't just alter consciousness. They can also affect the brain's overall ability to process information, and thus researchers might inadvertently confuse the absence of conscious experience with an impairment or change in cognitive processing itself.

In simpler terms, a researcher might think the experience measures what a person actually feels, when in fact it only partially measures the brain's ability to receive, process, or respond to the signal. This confusion becomes even more sensitive when similar indicators are used to make judgments about the consciousness of beings or systems that cannot describe their own experience. The study warns that this

methodological

problem could lead to claims that are more powerful than the evidence allows. In recent years, there has been much discussion about animal consciousness, the potential for consciousness in artificial intelligence, experiments with embryos, and brain organoids grown in laboratories for research purposes.

The study doesn't say that these entities are conscious or unconscious, but the evidence used in such discussions may be weaker than it appears if the indicators measure information processing rather than subjective experience. Therefore, the researchers call for stricter scientific standards before using the results of consciousness research in ethical or regulatory matters.