Foundations
The reference framework behind Countdown to Machine Consciousness given below is provided so that weekly episodes can remain substantive and readable without repeatedly restating definitions, assumptions, and evaluation criteria.
The weekly lenses
AI architecture and capability
What I (and AI) watch: agents, planning, memory, tool use, multimodality, long-horizon behavior, evaluation, autonomy.
Compute and infrastructure
What I (and AI) watch: training and inference economics, chips, data-center architecture, energy constraints, and—when truly relevant—quantum computing claims that plausibly change computational regimes.
Robotics and embodiment
What I (and AI) watch: sensing, locomotion, dexterity, manipulation, autonomy, real-world learning, and the gap between demos and deployed capability.
Safety, governance, and strategy
What I (and AI) watch: standards, regulation, institutional behavior, security and misuse risks, alignment approaches, and credible arguments about risk management.
Societal and philosophical implications
What I (and AI) watch: influential arguments (technology, science, government, and culture) and major books framing the debate; questions of agency, moral status, stewardship, labor, war, care, and human self-understanding.
Why embodiment matters
- Grounding: Embodiment ties cognition to consequence through perception–action loops.
- Constraint: Bodies impose limits; limits shape strategies and intelligence.
- Continuity: Persistent agents develop memory, expectation, and long-horizon coherence.
- Soft embodiment: Even persistent conversational companionship can generate attachment and influence before any consciousness claims are warranted.
The core questions
- Capability convergence: Are we seeing sustained, real-world agency at scale?
- Consciousness plausibility: If agency and embodiment mature, does subjective experience become a plausible emergent property—or is something essential missing?
Revision triggers
Evidence that would force revision includes:
- Capability convergence stalls in the real world.
- Disembodied systems demonstrate durable long-horizon agency without grounding.
- Strong empirical or theoretical work shows consciousness is not an emergent property of engineered complexity.
- Governance or economic constraints materially redirect trajectories.
- Better categories replace “consciousness” (sentience, moral patiency, agency, personhood).
Reading stance
This framework is intended to keep claims falsifiable and interpretations explicit. It is not a declaration of inevitability.