Consciousness Researchers Say Brain-as-Computer Metaphor Misleads AI Debate
Neuroscientists Anil Seth, Ned Block and Giulio Tononi argue in a Nature piece that today's AI systems are not conscious and are unlikely to become conscious even on silicon hardware.
Three leading consciousness researchers have warned in a commentary for the journal Nature that the idea of the brain as a computer has shaped the development of artificial intelligence and at the same time distorts the debate about consciousness. The authors are Anil Seth, Ned Block and Giulio Tononi. Their central thesis is that current AI systems are not conscious and that they will not develop consciousness even on silicon hardware.
At the heart of the criticism is the assumption that consciousness arises through computation alone. The authors describe this view as fundamentally flawed. For decades it has become an unspoken foundation of many considerations about machine intelligence. The metaphor of the brain as a computer has shaped thinking about AI and may obscure the view of the real question: what consciousness is and what it arises from. For this reason, the researchers see the metaphor as a factor that not only accompanies the debate but actively steers it in a particular direction.
From this position, the authors draw a practical consequence. They warn against granting rights to AI systems as long as it remains unclear whether such systems have any experience at all. The contribution was picked up and placed in context by the AI news portal The Decoder, which emphasized above all how strongly the computer conception has shaped expectations of machines.
At the same time, the researchers acknowledge that important questions remain open. According to their account, it is not settled whether any future system could be conscious. It is likewise unresolved which alternative theory of consciousness might be correct. Among scientists, the debate as a whole remains unsolved. The critique of the computer metaphor is therefore a position within this dispute and not a final answer.
For the discussion about AI, the contribution is significant insofar as it challenges a widespread way of thinking. If the equation of brain and computer shapes expectations of machines, it also influences which questions are asked and which answers are considered plausible. The authors therefore argue for openly examining the basic assumptions about the relationship between computation and consciousness before drawing far-reaching conclusions about future AI. This includes the question of whether a debate about rights for AI systems can meaningfully be conducted while the foundations of consciousness itself are disputed.
The researchers thus emphasize less a certainty than a caution in the handling of concepts. Anyone who wants to assess AI must, in their reading, first clarify what is understood by consciousness, and must not tacitly equate this question with the question of computing power.
Source: the-decoder.com
18
