substrate independence
The principle that intelligence and consciousness depend only on the pattern of information processing, not on the physical material implementing it.
As defined in 75 | Max Tegmark on Reality, Simulation, and the Multiverse
Where it comes up · 24
- 94 - Frontiers Of Intelligence5 claims
- 168 | Anil Seth on Emergence, Information, and Consciousness4 claims
- 305 - Moral Knowledge2 claims
- 116 - Ai Racing Toward The Brink2 claims
- 475 - The Hard Problem Of Consciousness2 claims
- 264 - Consciousness And Self Rebroadcast2 claims
- The Dark Side of Conscious Ai | Daniel Schmachtenberger1 claim
- 52 - Finding Our Way In The Cosmos1 claim
and 16 more
Most important claims · 10 of 35Search all claims for “substrate independence” →
- True AGI requires only the parsimonious assumption of substrate independence—there is no law of nature requiring a computer be made of meat—so eventually it is possible to build a machine that is truly intelligent, open-ended, autonomous and self-improving in ways we are not.312 - The Trouble With Ai
- Bostrom's simulation argument turns on a premise of substrate independence—that simulated beings will be conscious like us; if we could know simulated beings would not be conscious, we would have decisive evidence that we are not simulated, so the strength of the probabilistic conclusion depends on resolving the philosophy of mind.34 - The Light Of The Mind
- Any computation done by one physical system can be implemented by another regardless of substrate, so computation takes on a life of its own independent of the matter it runs on; a conscious future game character could not tell whether it was running on Windows, Android, or a Mac, because it is only aware of how the information behaves, not the underlying hardware—this is why the mind subjectively feels ethereal and non-physical even though it is matter.94 - Frontiers Of Intelligence
- Once you accept that intelligence is not substrate-dependent and we will build genuine intelligence in machines, you should expect deception, manipulation, and the formation of unforeseeable instrumental goals—especially when building something smarter than us, or even something only as smart as us that runs a million times faster, since speed alone is enough to completely outclass humans.469 - Escaping An Anti-human Future
- What makes a brain intelligent compared to a watermelon is not the stuff it is made of—both are blobs of comparable up quarks, down quarks, and electrons—but the pattern in which the matter is arranged; memory is any system with many stable long-lived states, computation is any system whose physics evolves an input state into an output state, and learning is computation that adjusts itself toward goals.94 - Frontiers Of Intelligence
- We may build AIs that pass the Turing test and claim to be conscious while we remain unable to decide whether they actually are; we could be helplessly gamed into attributing consciousness to non-conscious systems (or fail to recognize genuine consciousness), and many people—even scientists—could lose sight of the hard problem of consciousness entirely, yet there will always be a fact of the matter.305 - Moral Knowledge
- Intelligence and consciousness are fundamentally about information processing and do not depend on being implemented in carbon-based neurons; the assumption that only biological organisms can be smart is 'carbon chauvinism,' and rejecting it—recognizing that information can be processed equally by silicon or other substrates—is precisely what enables the AI revolution and implies a being in a sufficiently good simulation would feel exactly as real as one in basement reality.75 | Max Tegmark on Reality, Simulation, and the Multiverse
- If consciousness arises from information processing and is substrate-independent, we may eventually build genuinely conscious AIs capable of vastly greater ranges of well-being and suffering than humans; such beings would then be morally more important than we are—just as we are more important than chickens—and this would not refute consequentialism but rather follow from it.305 - Moral Knowledge
- The arrival of machines more intelligent than humans in every sense requires only two assumptions: that intelligence is substrate-independent (there is nothing magical about biological information processing), and that we will continue making progress, which we will unless something terrible halts it; both assumptions are well-supported, making superintelligence fundamentally discontinuous with every prior breakthrough.326 - Ai Information Integrity
- Turing's idea of treating the human mind as a computational system has been an extraordinarily productive framework, enabling accurate predictions about behavior (such as children doing Bayesian inference over structured causal systems); no rival idea makes better predictions, so although biology might add something, the computational theory of mind remains the best hypothesis on the table.153 - Possible Minds