From: mk_thisisit
The relationship between quantum physics and consciousness is a complex topic, with some theories suggesting that consciousness is not computational and may require a new understanding of physics.
Consciousness and AI
The term “artificial intelligence” is considered misleading, as true intelligence requires consciousness [00:00:04]. Computers are based on computational mathematics, which is a limited part of mathematics [00:00:17]. It is believed that computers will never gain consciousness because they are computational machines [00:00:24], and consciousness is not computational [00:00:30].
When discussing artificial intelligence, it is noted that these machines do not know what they are doing [00:00:35]. The increasing power of computers has led people to lose sight of their fundamental nature [00:00:41].
The Misnomer of “Artificial Intelligence”
The name “artificial intelligence” is deemed incorrect because intelligence necessitates consciousness [00:02:01]. Current devices are not conscious and will not be until they can invent something themselves; they merely perform calculations [00:02:13]. When people speak of intelligence, they often imply something conscious [00:02:36].
A more accurate term for what is currently called AI might be “artificial cleverness” [00:14:09]. This distinction highlights that while AI can be very clever and perform tasks efficiently, it does not necessarily understand what it is doing [00:14:18].
Gödel’s Theorem and Uncomputability
Insights into the nature of consciousness and computation stem from studies in mathematical logic, including Turing machines and the concept of computability [00:03:12]. Gödel’s theorem is considered incredible because it indicates that there are things whose understanding goes beyond their use [00:03:28].
Gödel’s theorem demonstrates that one can construct a statement whose meaning is understood, but which cannot be proven based on a given set of accepted computational rules [00:06:02]. The ingenuity lies in the fact that the rules for proof can be entered into a computer, meaning they are computational principles [00:04:44].
Crucially, there are “uncomputable” things in mathematics that cannot be entered into a computer [00:04:57].
- Computability means that a computing machine can be created to perform a task [00:09:09]. Different definitions of a computer (Turing, Church, Curry) have all been shown to be equivalent [00:05:17].
- AI cannot create its own rules because it doesn’t “know” that they are true [00:07:13]. The essence of Gödel’s theorem is about going beyond the rules by understanding why they are true [00:07:19].
- Understanding requires being aware of what one is doing and why the rules lead to truth [00:07:50].
- Consciousness allows for the transcendence of rules by understanding their truth [00:08:25].
The Nature of AI and Consciousness
AI’s role is primarily associated with computational power [00:10:17]. Computers can analyze vast amounts of data and identify patterns [00:10:30], but they do so without understanding their operation [00:10:45]. Understanding is distinct from calculating [00:10:49]. AI is fundamentally a statistical machine [00:10:57].
It is posited that consciousness is not computational [00:18:19]. This suggests that the answer to consciousness lies much deeper than current computational models [00:18:33].
The “Black Box” of Consciousness
As a physicist, the belief is that whatever creates consciousness belongs to the physical world, not magic [00:11:28]. The term “black box” for AI refers to knowing its inputs and outputs but having no idea what is inside [00:11:47].
It is claimed that the physics related to conscious thinking cannot be calculated [00:12:12], unlike much of known physics (e.g., general relativity, black holes) which is computable [00:12:36]. Therefore, artificial intelligence, being computational, will not be able to achieve consciousness [00:13:04].
Uncomputability in Nature
Computers are based on computational mathematics, which is a very limited section of the broader field of mathematics [00:15:59]. Much of mathematics goes far beyond computability and cannot be calculated using an algorithm [00:16:17].
The physics of our world can also be uncomputable [00:17:12]. Nature, through natural selection, may have created creatures that utilize this uncomputable aspect [00:17:22]. While humans are definitely conscious, the extent of consciousness in the animal kingdom (e.g., dogs, cats, elephants, octopuses, amoebas) is speculative [00:17:37].
Quantum Physics and Consciousness
The answer to consciousness is connected with a type of physics that is probably not yet known [00:18:41]. It lies deeper than the quantum world [00:18:51].
Computability in the Quantum World
The quantum world itself is still considered computable [00:19:00]. The uncomputable element concerns a fundamental unknown within the quantum world [00:19:03]. Quantum theory is seen as fundamentally incomplete and potentially wrong because it assumes quantum superpositions hold at all levels [00:19:08].
While it is possible that a universal quantum computer could create synthetic consciousness, this is a theoretical concept [00:19:23].
Classical vs. Quantum Reality
Two concepts of reality are distinguished:
- Classical reality: Can be determined and understood with our senses [00:24:36]. One can ask a classical system a question and receive a definitive answer about its state (e.g., the shape of a glass) [00:24:52].
- Quantum reality: Cannot be determined; it can only be confirmed [00:26:13]. For example, with the half-spin of a particle, one cannot ask its direction of rotation [00:25:29]. Einstein’s criterion suggested that if you can definitely confirm the existence of something, you receive an element of quantum [00:26:51].
Backward Causality and the EPR Paradox
Quantum reality behaves peculiarly, operating on the principle of backward causality [00:27:17]. This is reminiscent of experiments related to the EPR (Einstein, Podolsky, Rosen) paradox, where effects can influence what is momentarily present [00:27:31].
In the EPR paradox, if the initial spin of two entangled particles (Alice and Bob’s) is zero and their spins are opposite, Alice’s measurement of her particle’s spin retroactively determines Bob’s particle’s spin, but this influence travels backward along the past light cone, not simultaneously or faster than light [00:27:43]. Bob cannot determine his particle’s spin; he can only confirm it [00:28:24].
Understanding such phenomena requires the concept of quantum reality [00:28:49]. While signals cannot be sent faster than light, influence can travel backward or almost as fast as light, but only in quantum reality [00:28:59].
Understanding Quantum Mechanics
There is a famous quote attributed to Bohr, “Nobody understands quantum mechanics” [00:29:58]. This suggests that understanding in the classical sense may not apply to the quantum world [00:30:38]. The quantum world is non-deterministic and possibly incomprehensible to human minds in a classical way [00:30:24].
It is possible to know how to make the predictions of quantum mechanics work, even if they are probabilistic [00:30:56]. The key is to separate quantum reality from classical reality, acknowledging that quantum reality behaves differently [00:31:50].
Future of AI and Consciousness
AI machines will likely become better than humans in many tasks, even today [00:23:01]. However, this does not indicate consciousness [00:22:03]. The greatest risk in AI development is people mistakenly believing that machines are conscious [00:23:18]. The risk would be even greater if machines truly possessed consciousness [00:23:30].
If someone could understand what consciousness is and build a device based on it, it would be a different, conscious device, not a mere calculating machine [00:23:36]. Since consciousness is believed to be a physical process, its understanding may be connected to the mysterious collapse of the wave function and backward causality [00:23:49].