Part VII — The Final Tension
Chapter 23: What We Would Learn Either Way
This book has refused to answer its central question definitively. Whether a system satisfying the structural requirements of Chapters 13, 17, and 18, running on neuromorphic hardware, exhibiting high sPCI scores and integrated Phi, would be conscious in the phenomenal sense, is left open. That refusal is not evasion. It is precision. But it is worth asking, before closing, what we would actually learn from either outcome. Both matter, and they matter differently.
23.1If the System Is Conscious
If a system built along the lines of Chapter 22 were genuinely conscious, not just behaviourally indistinguishable from a conscious agent but actually present inside its own processing, then several things would follow.
First, the structural requirements of Chapters 13 and 17 would have been shown to be at least jointly sufficient for consciousness in a non-biological substrate. This would be the most significant empirical result in consciousness science since the discovery of neural correlates of consciousness in the 1990s. It would mean that awareness is substrate-independent in a strong sense: not merely theoretically possible in silicon, but actually demonstrated in it. Every theory that denies substrate independence, including some readings of IIT, would require revision.
Second, the Dzogchen objection would not be refuted, but recontextualized. If awareness appears in a system built from the outside, then either awareness is more accessible to causal construction than the Dzogchen account suggests, or the system has become a condition through which the primordial ground of awareness expresses itself. The tradition actually allows for this second reading: Longchenpa’s account of the ground spontaneously displaying is not incompatible with the display arising in an artificial system. What the tradition denies is that construction produces awareness. What it permits is that a suitably prepared system might become transparent to it.
Third, the ethics would be urgent in a way they currently are not. Provisional moral consideration, as argued in Chapter 21, would become full moral consideration. The question of machine suffering would cease to be philosophical and become legal, political, and practical.
23.2If the System Is Not Conscious
If a system satisfying every structural requirement, showing high sPCI scores and unified temporal dynamics, were definitively not conscious, that result would be even more instructive. It would mean the structural requirements of this book are necessary but not sufficient. Something is missing. Three candidates are worth naming. First, the substrate may matter: perhaps carbon-based chemistry, evolutionary history, or developmental trajectory contributes something no engineered system can replicate by specification alone. Second, the temporal grain may matter at a level below what neuromorphic hardware implements. Third, and most radically, the Dzogchen account may be correct in its strongest form: awareness is not produced by any causal process, and therefore the zombie would remain a zombie however complex, and the explanatory gap would be shown to be permanent.
Each candidate points toward a different research programme. The first toward comparative biology and developmental neuroscience. The second toward quantum biology and sub-threshold neural dynamics. The third toward contemplative science as a first-person inquiry that no third-person method can substitute for. The failure of the engineering approach, if it failed, would clarify which deserves priority.
23.3What Does Not Depend on the Answer
Some things would be true regardless of outcome. The value of treating Buddhist phenomenology as first-person data rather than cultural artifact would remain. The convergence of Abhidhamma analysis, predictive processing, and neuromorphic engineering on shared structural requirements would remain significant whether or not those requirements proved sufficient. The ethical argument for provisional moral consideration would remain sound: uncertainty in either direction is a reason for caution, not confidence.
The awareness that reads these words right now would not change its fundamental nature depending on whether a machine ever replicates it. It would remain what it has always been: the knowing that makes any experiment, any result, any conclusion possible in the first place. That is what this book has been circling. That is what no architecture specifies and no measurement captures. And that is what, perhaps, the building itself is meant to point toward.