Short Overview: Is Artificial Consciousness Achievable? Lessons from the Human Brain

The paper “Is Artificial Consciousness Achievable? Lessons from the Human Brain” by Michele Farisco, Kathinka Evers, and Jean-Pierre Changeux offers a rigorous evolutionary and neuroscientific examination of the challenges and pathways to developing artificial consciousness.

Key Highlights

• Human Brain as Benchmark:
  The authors propose that instead of theorizing consciousness in the abstract, efforts in artificial consciousness should use the evolution of the human brain and its layered architecture as a benchmark.

• Architectural Requirements:
  Complex features like hierarchical modularity, recursive feedback loops, and neuromodulation (e.g. dopamine systems) are emphasized as preconditions for achieving human-like conscious processing.

• Limitations of Current AI:
  The paper outlines both intrinsic (architectural) and extrinsic (scientific limitations) constraints preventing current AI from achieving true consciousness.

• Spectrum of Consciousness:
  It allows for the possibility of alternative forms of consciousness. These may be non-human and perhaps not even comparable. They could be developed through synthetic means.

• Modularity and Feedback Loops:
  ACM is structured to integrate recurrent feedback mechanisms and modular subsystems for learning and memory. These parallel the neural systems discussed in the paper.

1. Ethical Awareness:
   The ACM initiative shares the authors’ caution about claims of artificial consciousness, and advocates for transparent, open-source, and collaborative development with an ethical foundation.

However, while this paper leans toward a brain-centric approach to artificial consciousness, ACM remains open to both brain-inspired and alternative frameworks for generating emergent agency and awareness.

To read the full paper, visit:
https://www.sciencedirect.com/science/article/pii/S0893608024006385