If consciousness cannot be fully explained by physical processing, it may not be beyond science but may require something like another fundamental law or principle (called 'psychophysical laws') to connect physical processes to consciousness, analogous to fundamental laws in physics.

The history of science and philosophy repeatedly shows humanity is not at the center of things and not unique: we are not the center of the universe, not separate from other animals, not created by God—and similarly, human consciousness is one way of being conscious among many possible forms, likely instantiable in different material substrates.

Consciousness is often considered the gateway to moral consideration: once we acknowledge a system is conscious and can suffer, we have moral obligations to consider it in our ethical calculus; this applies to AI, language models, and organoids—if we fail to recognize consciousness in systems we create, we risk moral catastrophe.

We have already developed language models that are very close to passing the traditional Turing test (indistinguishable from humans), and some personas generated by GPT-4 have already passed 5-minute Turing tests; in the past this would have been evidence of consciousness, though now we may resist drawing that conclusion.

Chalmers argues that what matters for consciousness is informational properties of the brain rather than biological properties, and he is looking for a beautiful mathematical equation connecting information, computation, and consciousness.

When constructing conscious AI, a reasonable strategy is to base it on architectures mimicking the human brain because humans are the one case known to produce consciousness.

Chalmers discusses a thought experiment about gradually replacing biological neurons with silicon chips that replicate their function: if such replacement could proceed from 1% to 100% while the replaced individual reports continuity of consciousness, this would suggest silicon can be conscious.

The analogy between the hard problem of life and hard problem of consciousness is imperfect because life's hard problem concerned objective functional properties (reproduction, adaptation, metabolism) that could in principle be mechanistically explained, whereas consciousness's hard problem concerns subjective phenomenal experience (first-person experience) which has no analog in the case of life except consciousness itself.

Tononi Maps Correlations, Not The Hard Problem

Tononi's Integrated Information Theory does not solve the hard problem; instead it starts by taking consciousness as given, derives phenomenological axioms, turns them into a mathematical measure (phi) of integrated information, and aims at a rigorous theory of which physical states correlate with which conscious states—leaving open why integration yields consciousness at all.

C. Elegans Map Without Working Simulation

We have a perfect map of all the connections between the 302 neurons of the worm C. elegans, yet we still cannot make a simulation work because we don't understand the principles of how the components operate well enough—illustrating that connectivity maps alone are insufficient for functional simulation.

Recursive Self-Improvement Leads To Intelligence Explosion

Once machines become a bit smarter than us, they will be a bit better at designing machines, which will design machines smarter still, and this process continues recursively until fairly soon you have machines vastly smarter than we are—an informal argument that can be made fairly rigorous, making such a singularity pretty likely unless fairly strong conditions hold to prevent it.

Simulation Speculation Revives A Creator

Although a natural atheist about gods, thinking about simulations makes a creator more plausible—'simulation theology'—since living in a simulated world implies there could be a creator of this local bit of the universe, perhaps merely a teenage hacker in the universe above, an idea even a naturalist can take somewhat seriously.

Consciousness Is Subjective First-Person Experience

Consciousness is the subjective experience of the mind and world—what it feels like from the first-person point of view to think, perceive, and judge—which must be distinguished from the war of information processing and behavior that accompanies it.

The Easy Versus Hard Problems Of Consciousness

The 'easy problems' of consciousness concern explaining functions—how we discriminate information, integrate it, control behavior, and monitor our own states—which fall within standard cognitive science by finding mechanisms; the 'hard problem' is why all that processing feels like something from the inside, which is not a problem about behavior or function and therefore resists the usual mechanistic methods.

Consciousness Studies Has Wide Disagreement

There is a huge amount of disagreement in the field, but most people have at least a reasonable appreciation that there is a big problem here, even if what they do after recognizing it differs greatly.

Dennett Relabeled Eliminativism As Deflated Consciousness

Dennett's view evolved from saying consciousness doesn't exist (no phenomenology, quining qualia) to saying consciousness exists only in a deflated sense of functioning, behavior, and encoded information—which is ultimately just a verbal relabeling that still denies consciousness in the strong subjective sense that poses the hard problem, analogous to compatibilist relabeling in the free will debate.

Epiphenomenalism As The Counterintuitive Default

Because nearly any behavior can in principle be explained by neural or computational mechanisms without invoking subjective consciousness, one is pushed toward epiphenomenalism—the view that consciousness, like smoke from a locomotive's smokestack, accompanies but does not cause physical events—even though no one finds it initially attractive and it raises puzzles about how consciousness evolved and why we even talk about it.

The Zombie Argument Throws Function Into Relief

We can conceive of a 'philosophical zombie'—a being physically identical to a normal human but lacking any subjective consciousness—and the very conceivability of such a being raises the question of why we aren't zombies, since for any behavioral function (perception, learning, memory, language) it appears a zombie could perform it unconsciously, throwing the question of what consciousness does into harsh relief.

The Zombie Talking About Consciousness Puzzle

A deep puzzle in the zombie thought experiment is why a zombie devoid of phenomenal experience would talk about consciousness at all—but since there are brain mechanisms responsible for everything we say and do, those same mechanisms would produce the zombie's reports, raising the possibility that the things we say about consciousness (an easy problem) might be explicable in physical terms even if the experience itself is not.

IIT Exclusion Axiom Seems Arbitrary

Tononi's theory implies that highly integrated subsystems of the brain (like a hemisphere) would themselves be conscious, forcing him to add an 'exclusion axiom' stating that if you are part of a system with higher phi than you, you are not conscious—an axiom many find arbitrary, since without it you would be left with conscious subsystems all over.