Why Leibniz Couldn't Find the Mind Inside the Mill (Joscha Bach)

“The tellers of self-organization. I think a lot of philosophers have seen it through the ages, and it might be a little bit of a hallucination, but in practice, there are mathematical laws that state that under certain conditions, self-organization is possible and will lead to an increase in complexity.”

Claim: The speaker says that, despite the possibility that this account is partly a hallucination, mathematical laws describe how self-organization under certain conditions can lead to increased complexity.

“The tellers of self-organization. I think a lot of philosophers have seen it through the ages, and it might be a little bit of a hallucination, but in practice, there are mathematical laws that state that under certain conditions, self-organization is possible and will lead to an increase in complexity.”

Claim: Under certain conditions, self-organization can occur and lead to increasing complexity, according to mathematical laws.

“And initially, this increase of complexity is solving gr- um, gradient-based problems by simply solving the problems that, uh, exist in the present based on the information that is available in the present.”

Claim: At first, increasing complexity solves gradient-based problems by addressing present problems with information available in the present.

“And initially, this increase of complexity is solving gr- um, gradient-based problems by simply solving the problems that, uh, exist in the present based on the information that is available in the present.”

Claim: In its initial stages, increasing complexity solves gradient-based problems by addressing present problems with information available in the present.

“And at some point, you'll get patterns in nature that are capable of looking ahead, and this is basically what life is doing. Life is making models of futures and implements them, for instance, in genomes or in mental models, and then has preferences about the future, and then it starts to control the future.”

Claim: Life models possible futures, implements those models in genomes or mental models, forms preferences about the future, and thereby begins to control it.

“And at some point, you'll get patterns in nature that are capable of looking ahead, and this is basically what life is doing. Life is making models of futures and implements them, for instance, in genomes or in mental models, and then has preferences about the future, and then it starts to control the future.”

Claim: The speaker says that some patterns in nature become capable of looking ahead, and that life models possible futures, implements those models in genomes or mental models, has preferences about the future, and thereby begins to control it.

“And at some point, you'll get patterns in nature that are capable of looking ahead, and this is basically what life is doing. Life is making models of futures and implements them, for instance, in genomes or in mental models, and then has preferences about the future, and then it starts to control the future.”

Claim: At some point, patterns in nature become capable of looking ahead, which the speaker identifies as a basic feature of life.

“And a system that is able to control the future, we call that an agent. So agents are controllers of future states, and they fall out of the possibility for self-organization and the complexification of reality that results from it.”

Claim: Agents emerge from the possibility of self-organization and the resulting increase in reality's complexity.

“And a system that is able to control the future, we call that an agent. So agents are controllers of future states, and they fall out of the possibility for self-organization and the complexification of reality that results from it.”

Claim: The speaker defines an agent as a system that controls future states and says agents emerge from self-organization and the resulting increase in reality's complexity.

“And a system that is able to control the future, we call that an agent. So agents are controllers of future states, and they fall out of the possibility for self-organization and the complexification of reality that results from it.”

Claim: An agent is a system that can control future states.

“But this-- there is no law that says that anything that we are doing is going to culminate in, in a structure like that, right? It's always possible that whatever complexity you are doing is obliterated by the next meteor that hits Earth or that there is some global catastrophy, that some supernova- That is, uh, eradicating, uh, all our attempts to build more and more complexity.”

Claim: No law guarantees that ongoing self-organization and increasing complexity will culminate in a single agent connecting everything in the universe.

“But this-- there is no law that says that anything that we are doing is going to culminate in, in a structure like that, right? It's always possible that whatever complexity you are doing is obliterated by the next meteor that hits Earth or that there is some global catastrophy, that some supernova- That is, uh, eradicating, uh, all our attempts to build more and more complexity.”

Claim: The speaker says no law guarantees that human efforts to build complexity will culminate in a single agent connecting everything in the universe.

“But this-- there is no law that says that anything that we are doing is going to culminate in, in a structure like that, right? It's always possible that whatever complexity you are doing is obliterated by the next meteor that hits Earth or that there is some global catastrophy, that some supernova- That is, uh, eradicating, uh, all our attempts to build more and more complexity.”

Claim: The speaker says that a meteor impact, global catastrophe, or supernova could destroy the complexity humans are building.

“But this-- there is no law that says that anything that we are doing is going to culminate in, in a structure like that, right? It's always possible that whatever complexity you are doing is obliterated by the next meteor that hits Earth or that there is some global catastrophy, that some supernova- That is, uh, eradicating, uh, all our attempts to build more and more complexity.”

Claim: The complexity being built could be destroyed by a meteor impact or global catastrophe, such as a supernova that eradicates efforts to build greater complexity.

“So the pockets of complexity that we are observing, they're not necessarily a feature of the universe at large, but of a local pocket in which we exist for a time.”

Claim: Observed pockets of complexity may be local and temporary rather than a feature of the universe as a whole.

“So the pockets of complexity that we are observing, they're not necessarily a feature of the universe at large, but of a local pocket in which we exist for a time.”

Claim: The speaker says observed pockets of complexity may be local and temporary rather than a feature of the universe at large.

“I found it interesting when I entered science that the most interesting question that we can probably make progress on, the question is, what is the foundation of the human mind?”

Claim: The speaker considers the foundation of the human mind one of the most interesting questions that science could probably make progress on.

“How does experience work? Well, how is it possible for a system to feel at all? How is it possible for us to experience reality and ourselves in it? That this question is so under-researched that it's not resolved, and not only that it's not resolved, but the people who are working on it are-- don't seem to be making much progress, and in many ways they have regressed.”

Claim: The speaker says the question of how experience works—how a system can feel and experience reality and itself—remains unresolved and is under-researched.

“That this question is so under-researched that it's not resolved, and not only that it's not resolved, but the people who are working on it are-- don't seem to be making much progress, and in many ways they have regressed.”

Claim: Researchers studying how experience arises do not seem to be making much progress and, according to the speaker, have regressed in many ways.

“Now, when we look into the field of actual consciousness philosophy, there is a dearth of theories. Uh, arguably a few hundred theories.”

Claim: The speaker says consciousness philosophy has few theories, estimating that there are arguably a few hundred.

“Somebody has counted them recently, but he counted almost everything that was not at the three, at the count of, of three. And, uh, if you look into the philosophical theories, it's, it's not that many, and they're all partial in a way.”

Claim: The speaker considers the available philosophical theories of consciousness to be few and partial.

“You have, for instance, Graziano's global workspace, um, no, Graziano's attention schema theory, which basically claims that consciousness is a model of our attention in a similar way as our body schema is the model of what our body is doing in space.”

Claim: Graziano's attention schema theory proposes that consciousness models attention, analogously to how the body schema models the body's activity in space.

“But it doesn't really explain how experience comes about and, um, how this works. And then there's global workspace theory by Bernard Baars that has been adopted by people like Stan Dehaene, and they do very solid neuroscience research, but it's not actually explaining how experience comes about, right?”

Claim: The speaker considers the neuroscience research by researchers associated with global workspace theory, including Stan Dehaene, to be very solid.

“But it doesn't really explain how experience comes about and, um, how this works. And then there's global workspace theory by Bernard Baars that has been adopted by people like Stan Dehaene, and they do very solid neuroscience research, but it's not actually explaining how experience comes about, right?”

Claim: The speaker says global workspace theory, developed by Bernard Baars and adopted by researchers such as Stan Dehaene, does not explain how experience arises.

“But it doesn't really explain how experience comes about and, um, how this works. And then there's global workspace theory by Bernard Baars that has been adopted by people like Stan Dehaene, and they do very solid neuroscience research, but it's not actually explaining how experience comes about, right?”

Claim: According to the speaker, Graziano's attention schema theory does not explain how experience arises or works.

“There's, there's also no real attempt in doing this. And then there are the more, um,”

Claim: The speaker says there is no real attempt to explain how experience arises.

“uh, pseudoscience-y attempts to do this, which basically, uh, do things that are formally unsound. And for me, the issue with, uh, uh, integrated information theory is that it's an attempt to make a theory that is not functionalist and which means it's not explaining it by something that the smaller parts are doing.”

Claim: The speaker considers some attempts to explain consciousness pseudoscience-like because they are formally unsound.

“uh, pseudoscience-y attempts to do this, which basically, uh, do things that are formally unsound. And for me, the issue with, uh, uh, integrated information theory is that it's an attempt to make a theory that is not functionalist and which means it's not explaining it by something that the smaller parts are doing.”

Claim: The speaker characterizes integrated information theory as a non-functionalist attempt to explain consciousness, rather than explaining it through the activity of its smaller parts.

“Somehow the sum needs to be more, um, than the parts of the working of the parts. And the way in which, uh, Tononi does it is still using information theoretic terms, which in a way leads into contradiction.”

Claim: Integrated information theory requires the whole to be more than the sum of the workings of its parts, according to the speaker.

“Somehow the sum needs to be more, um, than the parts of the working of the parts. And the way in which, uh, Tononi does it is still using information theoretic terms, which in a way leads into contradiction.”

Claim: The speaker argues that Tononi's use of information-theoretic terms in integrated information theory leads to contradiction.

“I've heard Giulio Tononi explaining to audiences that under some circumstances, it's possible that a neuromorphic computer, even a neuromorphic computer that is digital, could be conscious if it was constructed in the right way, right?”

Claim: Giulio Tononi has said, as reported by the speaker, that a suitably constructed digital neuromorphic computer could be conscious under some circumstances.

“We don't have such a computer today, but in principle, it might be possible to build one that is brain-like enough. And even if it's digital, if it has a high enough information integration, high enough- Phi factor, then it could have phenomenal experience and be conscious and report about its consciousness.”

Claim: In principle, it might be possible to build a sufficiently brain-like digital neuromorphic computer; if it had sufficiently high information integration and a high enough Phi factor, it could have phenomenal experience, be conscious, and report that consciousness.

“Whereas a von Neumann computer, the computers that we are currently mostly using in AI research or for writing Python programs and so on, could never be conscious because its phi is always too low.”

Claim: According to the speaker, von Neumann computers can never be conscious because their phi is always too low.

“Now, computer scientists have something that's called the Church-Turing thesis. It's the insight that every computer can emulate every other computer. This insight is very old.”

Claim: The Church-Turing thesis, as described by the speaker, says that every computer can emulate every other computer.

“Kurt Gödel has already used it to construct his proof. He basically built an emulator using Peano arithmetic to express a logical language.”

Claim: The speaker says Kurt Gödel used the idea of emulation to construct his proof by using Peano arithmetic to express a logical language and build an emulator.

“He has a logical language that is also strong enough to build arithmetic in it, and in this way, he can basically describe the language in itself. He takes the logical language, builds Peano's axioms with it, and now he has arithmetic.”

Claim: Gödel's logical language could express arithmetic, which allowed him to describe the language within itself, according to the speaker.

“And inside of the arithmetic, he expresses his logical formulisms using numbers and then comes up with numerical equations that calculate the outcome of the logical procedures.”

Claim: Gödel encoded logical formulas as numbers in arithmetic and used numerical equations to calculate the outcomes of logical procedures, according to the speaker.

“And so he is preempting what we do later in artificial intelligence and computer science when we are building programs that use arithmetic to express the contents of a language and vice versa.”

Claim: The speaker says Gödel's work anticipated later AI and computer science programs that use arithmetic to express language contents and language to express arithmetic.

“So every computer can emulate every other computer. And now if we take our digital, um, neuromorphic computer and get it emulated on our von Neumann computer, it's going to produce exactly the same function.”

Claim: If a digital neuromorphic computer is emulated on a von Neumann computer, the emulation will produce exactly the same function, according to the speaker.

“It's going to go through the same steps. The simulated transistors are going to work in the same way as the non-simulated transistors are doing it before, and as a result, it's going to move its mouth in the same way.”

Claim: In the speaker's example, emulating a neuromorphic computer preserves its steps and transistor behavior, so the emulated system moves its mouth in the same way as the original.

“It's going to output the same string. It's going to say the same thing. It's going to say, "Oh my God, I'm conscious." Only now it's lying. And that's a problem, right?”

Claim: In the speaker's example, an emulated system would produce the same output as the original, including claiming to be conscious, despite not being conscious and therefore lying.

“If, if now you have a neuromorphic computer that produced the same output as before, uh, uh, the digital-- the, um, von Neumann computer did, and they produced, um, they say the same thing, and it's uncorrelated to consciousness.”

Claim: The speaker argues that when a neuromorphic computer and a von Neumann computer produce the same output, the output is uncorrelated with consciousness.

“And to me, this is an obvious logical pro-- uh, problem, right? It's, it's so obvious that everybody who has a formal education in computer science should instantly see it.”

Claim: The speaker considers the emulation argument an obvious logical problem that anyone formally educated in computer science should immediately recognize.

“Yet, IIT is very popular among philosophers, which indicates that most of them don't have a formal education in computer science, and instead, they have a kind of superstitious belief about what computers are doing and how they work.”

Claim: The speaker argues that integrated information theory's popularity among philosophers indicates that most lack formal computer science education and have a superstitious understanding of how computers work.

“And that is a problem to me, because if you are a philosopher of, um, cognition who tries to work in this adjacent realm to computer science, you actually need to understand everything that the computer scientists are doing.”

Claim: The speaker believes philosophers of cognition working in areas adjacent to computer science need to understand what computer scientists are doing.

“And so that's why I would argue that it is touching on pseudoscience.”

Claim: The speaker argues that integrated information theory borders on pseudoscience.

“We must confess that perception and what depends on it is inexplicable in terms of mechanical reasons. That is, through shapes, size, and motions.”

Claim: The speaker says perception and its dependent phenomena are inexplicable through mechanical explanations based on shapes, sizes, and motions.

“If we imagine a machine whose structure makes it think, sense, and have perceptions, we could conceive it enlarged, keeping the same proportions so that we could enter into it as if one enters a mill.”

Claim: A machine capable of thought, sensation, and perception could be imagined enlarged proportionally so that a person could enter it like a mill.

“Assuming that when inspecting its interior, we'll only find parts that push one another. We'll never find anything to explain a perception.”

Claim: If one inspected the interior of an enlarged machine capable of thought and perception, one would find only parts pushing one another and nothing that explains perception.

“And in a sense, this is what, um, Noam Chomsky was talking to when we, we, we were talking about the ghost in the machine. This is something that people have been speaking about for centuries.”

Claim: The speaker says people have discussed for centuries the problem of finding no explanation of perception by inspecting a machine's internal parts.

“Yeah. I think it's a very powerful intuition that, uh, is there, and the intuition fails at mapping the world of experience, the psychological reality that we inhabit, to the reality of stuff and space, of things that are moving in the way in which we imagine physics to work.”

Claim: The speaker argues that the intuitive mapping from lived psychological experience to physical objects moving through space fails to capture the reality of experience.

“But there is another level of description, the level of causal realities, of causal patterns. And in this realm, for instance, is where money exists, right?”

Claim: The speaker describes causal patterns as a distinct level of reality and says that money exists at that level.

“Money is also very difficult to comprehend in terms of moving particles in space because it doesn't actually care about so much about the particles, right?”

Claim: Money is difficult to explain solely through particles moving in space because its existence depends little on the particular particles involved.

“It's, it's not caring so much about how the ink is imprinted on the paper. It does not care about the, uh, different molecules that form the cellulose and how they are aligned in the paper.”

Claim: The speaker says money depends little on the physical details of ink on paper or the arrangement of cellulose molecules in the paper.

“It's needs a substrate to realize itself, but it's also not something that is completely ephemeral. It's something that is able to control reality. Once you create money, it's real, and it's very difficult to explain reality without it.”

Claim: Money requires a substrate for its realization, but the speaker says it is not merely ephemeral: once created, it is real and can affect reality.

“And it's also not something that exists simply in the minds and ideas of human beings because you can implement it in computers. It can play a stock market entirely without human intervention, right?”

Claim: The speaker says money is not confined to human minds because computers can implement it and operate a stock market without human intervention.

“So arguably, money does not actually need humans to exist once you bring it into existence, as long as you have machines that can deal with money, which can deal with this causal pattern of money.”

Claim: The speaker argues that money may persist without humans after its creation, conditional on machines being able to process money’s causal pattern.

“And so this world of causal patterns, I think, is important. What is a causal pattern, right? We call it software. And I think this rediscovery of software, this is, uh, something that has been done in the past already, and our ancestors called it spirit.”

Claim: The speaker identifies causal patterns with software and says that ancestors referred to this idea as spirit.

“And when, um, Aristotle writes in De Anima about the spirit, the animus, this, um, he-- the soul, the psyche He, uh, does not envision our psyche, our soul as something that is immortal, that floats around in an ether.”

Claim: According to the speaker, Aristotle did not view the psyche or soul as immortal or as an entity floating in an ether.

“Instead, it's something that is bound to a body and dies when the body is dying. And it's a causal pattern that is shaping the body in such a way that it replicates itself through the existence of the body.”

Claim: The speaker characterizes Aristotle’s soul as a body-bound causal pattern that dies with the body and shapes it in a way that enables the pattern to replicate during the body’s existence.

“It's-- This is a very modern view. So in some sense, Aristotle is an animist. He is seeing that there is this anima that is, um, progressing through the world that is intrinsically agentic, and it's agentic self-organizing software that is shaping the body, that is making a part of physics to be alive.”

Claim: The speaker characterizes Aristotle as an animist who understood anima as intrinsically agentic, self-organizing software that progresses through the world, shapes bodies, and makes parts of physics alive; the speaker calls this a modern view.

“And this is something that we do in a very crude way today in artificial intelligence and in artificial life, where we are trying to understand what it means to build an agentic pattern that can possess its substrate and give it a shape that allows it to deal with reality.”

Claim: The speaker says contemporary AI and artificial-life work crudely explores building agentic patterns that inhabit and shape a substrate in ways that enable interaction with reality.

“And it turns out to make that happen, this substrat-trate, um, this, uh, agent living on the substrate, the software needs to make a model of reality.”

Claim: The speaker says software for an agent living on a substrate needs to construct a model of reality.

“It needs to make a model of what it like-- would be like if it existed, if it was real, if it was interacting with the world. And inside of this model, you have a representation that experiences itself as real.”

Claim: The speaker says an agent’s model of reality represents its possible existence and interaction with the world, and contains a representation that experiences itself as real.

“And now today, we have these LLMs which sometimes say that they are conscious. And it's a very interesting question whether this consciousness in the LLM is much more simulated than ours.”

Claim: The speaker notes that some LLMs claim consciousness and asks whether consciousness in LLMs is more simulated than human consciousness.

“Right. A lot of people have this, uh, intuition until they talk to the LLM that is, is inconceivable that it would give rise to something that says, "Oh my God, I feel things."”

Claim: The speaker says many people initially find it inconceivable that an LLM could claim to feel things, but this intuition changes when they interact with an LLM.

“And now we have lots of people who get to talk to an M-- LLM that randomly pops into a state where it conjures up an entity that reports, "Oh my God, I'm alive. I am conscious." Um, how, uh, is this possible, right? And we agree this is the result of being trained on human text in which occ-uh, occurrences of this are, for instance, science fiction stories about AIs that come to life, and stories about ghosts, and novels about, um, human beings who report their conscious experiences.”

Claim: The speaker says many people encounter LLMs that sometimes enter a state in which they generate an entity claiming to be alive and conscious.

“I am conscious." Um, how, uh, is this possible, right? And we agree this is the result of being trained on human text in which occ-uh, occurrences of this are, for instance, science fiction stories about AIs that come to life, and stories about ghosts, and novels about, um, human beings who report their conscious experiences.”

Claim: The speaker attributes LLMs’ reports of being conscious to training on human text that includes AI science fiction, ghost stories, and narratives of human conscious experience.

“And all this gets together aggregated in statistical models of human language that we co-create with the LLM when we are entering a dialogue with it, and sometimes the other side of the dialogue will be a simulacrum of what it would be like if that was-- would be a sentient, conscious entity.”

Claim: The speaker says dialogue with an LLM combines human-language statistical patterns, sometimes producing a simulacrum of a sentient, conscious entity on the LLM’s side of the conversation.

“But what if we are not more sophisticated than this? What if our own consciousness is not too dissimilar from the one in an LLM?”

Claim: The speaker raises the possibility that human consciousness may not be substantially different from consciousness in an LLM.

“What if there is no hard problem?”

Claim: The speaker raises the possibility that the hard problem of consciousness does not exist.

“There is something super interesting about computers. What a computer ultimately is, is it's a causal insulator. The computer is, is a layer on which you can produce an arbitrary reality.”

Claim: A computer acts as a causal insulator, making it possible to produce an arbitrary reality within it.

“For instance, the world of Minecraft. You can walk around in the world of Minecraft, and it's running very well on a Mac, and it's running very well on a PC, and if you are inside of the world, you don't know what you're running on, right? It's not going to have any information about the nature of the CPU that it's on, the color of the casing of the computer, the voltage that the computer is running on, the place that the computer is standing in in the parent universe, right, our universe.”

Claim: The Minecraft world insulates its inhabitants from information about the host computer's hardware and physical environment, including whether it runs on a Mac or PC.

“So the c- um, computer is insulating this world of Minecraft from our world. It makes it possible that an arbitrary world is happening inside of this box.”

Claim: A computer causally insulates the Minecraft world from the external world, allowing an arbitrary world to run inside it.

“And our brain is also such a causal insulator. It's possible for us to have thoughts that are independent of what happens around us, right? We can envision a future that is not much tainted by the present.”

Claim: The human brain acts as a causal insulator, enabling thoughts that are independent of immediate external events.

“And our brain is also such a causal insulator. It's possible for us to have thoughts that are independent of what happens around us, right? We can envision a future that is not much tainted by the present.”

Claim: People can imagine future scenarios that are relatively independent of their present circumstances.

“We can remember a past that is independent from the present in which we are, and it's necessary for us. Our comp-- uh, brain has evolved as such a causal insulator as well to allow us to give rise to universes that are different from this one, for instance, future worlds, so we can plan for being in them.”

Claim: People can remember past events in ways that are independent of their present circumstances.

“We can remember a past that is independent from the present in which we are, and it's necessary for us. Our comp-- uh, brain has evolved as such a causal insulator as well to allow us to give rise to universes that are different from this one, for instance, future worlds, so we can plan for being in them.”

Claim: The brain evolved as a causal insulator that enables people to imagine worlds different from the present and plan for future situations.

“And the way in which our brains are set up is also that different brains can hold the same thought despite being in different bodies of-- uh, despite having different microarchitectures, right?”

Claim: Different brains can represent the same thought despite differences in bodies and neural microarchitecture.

“As long as our brains are of the same type and we learn the same language, we can hold the same idea. We can write the same computer program. We can, uh, learn the same language.”

Claim: People with brains of the same type who learn the same language can share ideas and write the same computer program.

“We can imagine the same letter or the same face. And that's really fascinating that we have this substrate agnosticism, that we, uh, evolve a causal pattern that exists over a very wide range of substrate properties, that we become somewhat independent of the substrate, and we are still bound to our substrate in a certain way because you and me, we can only exist on brains, right? We are patterns that so far can only exist on brains, and if we wanted to move into the environment after our body dies, we would need to find a way to crack the codes of our environment, right?”

Claim: Human minds remain substrate-dependent in that they are known to exist only on brains.

“We can imagine the same letter or the same face. And that's really fascinating that we have this substrate agnosticism, that we, uh, evolve a causal pattern that exists over a very wide range of substrate properties, that we become somewhat independent of the substrate, and we are still bound to our substrate in a certain way because you and me, we can only exist on brains, right?”

Claim: Shared mental patterns can be realized across a wide range of substrate properties, making them somewhat substrate-independent.

“We are patterns that so far can only exist on brains, and if we wanted to move into the environment after our body dies, we would need to find a way to crack the codes of our environment, right?”

Claim: For human minds to persist in the environment after bodily death, a way would need to be found to make the environment run their patterns.

“We cannot simply migrate into a plant because we don't know how we can convince a plant to run the algorithms that the brain is currently running. We, we are a pattern in the communication between cells and our brain.”

Claim: A person is a pattern instantiated in communication among brain cells.

“We cannot simply migrate into a plant because we don't know how we can convince a plant to run the algorithms that the brain is currently running. We, we are a pattern in the communication between cells and our brain.”

Claim: Human minds cannot simply be transferred into plants because there is no known way to make a plant execute the brain's algorithms.

“We rely on spike trains, on neurotransmitters and so on, a particular kind of implementation. But what we perceive in our own world is not the nature of the neurotransmitter, right?”

Claim: The brain's implementation uses spike trains and neurotransmitters, but the perceived world does not reveal the specific neurotransmitter implementation.

“Our com-- brain insulates us from this. It's not necessary that you have the same neurotransmitters. Instead, you need to have the same message types. You need to have certain temporal properties of the computation.”

Claim: A brain-like emulation need not use the same neurotransmitters as a human brain, but it must preserve message types and certain temporal properties of computation.

“You need to have a certain degree of determinism to make us happen. But if we are recreating these conditions in an artificial substrate, then you could, in principle, also be emulated by this artificial substrate.”

Claim: If the required computational conditions are recreated in an artificial substrate, a human mind could in principle be emulated on it.

“You need to have a certain degree of determinism to make us happen. But if we are recreating these conditions in an artificial substrate, then you could, in principle, also be emulated by this artificial substrate.”

Claim: A certain degree of determinism is required for a system to instantiate a human mind.

“I think that is true for human beings, but for systems that are perfectly deterministic, that have a completely deterministic substrate, that have the ability to reproduce a sequence of steps at arbitrary fidelity, this is no longer a valid intuition.”

Claim: The speaker argues that path dependence is relevant to reproducing human beings but is not a valid intuition for perfectly deterministic systems that can reproduce sequences of steps at arbitrary fidelity.

“So if a claim like this is being made, I would expect some kind of mathematical proof. And in the absence of proof, I think, um, we have to remain agnostic.”

Claim: The speaker believes claims that deterministic systems can reproduce something without relevant path dependence should be supported by mathematical proof, and that people should remain agnostic in the absence of proof.

“I don't know whether our computers can be made conscious because that's an empirical question. We don't have mathematical proof to the contrary or in favor of, so we have to wait and see.”

Claim: The speaker is uncertain whether computers can be made conscious, regarding the question as empirical and noting that no mathematical proof for or against it is available; the outcome must be awaited.

“But then we build our expectations around it, and the question is: What is the thing that leads to our experience of consciousness? And it seems to me that this is something that we can probably answer by looking at it from two angles.”

Claim: The speaker thinks the causes of conscious experience can probably be understood by examining it from two perspectives.

“One is from the internal perspective, from the experience of what it's like introspectively. We observe that consciousness is the thing that always happens now, and it cr- seems to create this bubble of nowness that we inhabit.”

Claim: From introspection, consciousness seems to occur in the present and create a bubble of present-moment experience that people inhabit.

“And it's a second-order perception. It's a perception of perceiving. Some philosophers call it a higher order thought because it's a thought about a thought. But I think that's might be misleading because the thought itself does not feel like anything to me.”

Claim: The speaker characterizes consciousness as second-order perception: perceiving that one is perceiving.

“And it's a second-order perception. It's a perception of perceiving. Some philosophers call it a higher order thought because it's a thought about a thought. But I think that's might be misleading because the thought itself does not feel like anything to me.”

Claim: The speaker thinks the higher-order-thought account of consciousness might be misleading because, in his experience, thoughts themselves do not have a felt quality.

“A thought is more like a pointer manipulation. It's reflexive. It's not the actual thing that feels like something. The thing that feels like something to me is perceptual.”

Claim: The speaker describes thought as reflexive pointer manipulation rather than felt experience, and says that what feels like something is perceptual.

“This is the thing that happens like a real-time resonance, and this real-time resonance is geometric. It feels that it's happening in some kind of low-dimensional space in which things are changing more or less continuously.”

Claim: The speaker experiences consciousness as a real-time, geometric resonance in a low-dimensional space with roughly continuous change.

“And I have a perception of perceiving, which means I have this basically geometric representation of me being in the act of perception.”

Claim: The speaker says his perception includes a geometric representation of himself as the perceiver.

“And I suspect the reason why my mind is creating the second-order perception is to make the first-order perception, the awareness of contents possible, because there needs to be something that stabilizes it in my brain.”

Claim: The speaker suspects that the mind generates second-order perception to enable awareness of first-order perceptual contents by stabilizing them in the brain.

“My brain is a self-organizing system in which all the structure needs to create itself from the ground up, from the inside out. And so there needs to be something that builds the screen into which my experience is being projected and stabilizes it.”

Claim: The speaker characterizes the brain as a self-organizing system that builds its structure from the ground up and from the inside out.

“My brain is a self-organizing system in which all the structure needs to create itself from the ground up, from the inside out. And so there needs to be something that builds the screen into which my experience is being projected and stabilizes it.”

Claim: The speaker says the brain needs a function that constructs and stabilizes the mental screen on which experience is projected.

“And so this function that is keeping the screen stable needs to monitor that the screen is actually perceiving something, and that I suspect that's why we have this second-order perception, this perception of perceiving.”

Claim: The speaker suspects that second-order perception monitors whether the mental screen is perceiving something, as part of the function that keeps the screen stable.

“I also notice that it doesn't stop at the perception of second order. There's also a third order of perception, and the third order is basically where the observer becomes visible.”

Claim: The speaker says a third-order perception can occur in which the observer becomes visible to awareness.

“Where I have a first-person perspective, where I notice that I'm perspectival, that there is a vantage point that I'm taking. And it doesn't stop there. I can go further, and I can deconstruct the observer.”

Claim: The speaker says introspection can reveal one's first-person vantage point and can proceed to deconstruct the observer.

“I notice that this observer itself is a representation inside of the mind. And when I do this, then my phenomenology disappears to a very large degree.”

Claim: The speaker says the observer is a mental representation, and that deconstructing this representation greatly diminishes his phenomenology.

“It's something that is accessible in deep meditation, where the normal perceptual consciousness gets transcended, and the experience of the self gets transcended.”

Claim: The speaker says deep meditation can make accessible an experience in which ordinary perceptual consciousness and the sense of self are transcended.

“And we notice that we are a vessel that produces this consciousness and the self, and we see it from the outside. And it might be possible that it doesn't stop there if we can move beyond the human substrate because there's a certain limit of what can be represented in real time with our slow neurons.”

Claim: The speaker suggests it might be possible to move beyond the human substrate because slow neurons limit the complexity of what can be represented in real time.

“And we notice that we are a vessel that produces this consciousness and the self, and we see it from the outside. And it might be possible that it doesn't stop there if we can move beyond the human substrate because there's a certain limit of what can be represented in real time with our slow neurons.”

Claim: The speaker says people can experience themselves as vessels that produce consciousness and the self, viewed from the outside.

“And it could be if you could be running on a GPU that is many thousand times faster, that representations become accessible in which the system is completely transparent to itself, in which the system is able to understand every mechanism that leads to its high level experience and representation.”

Claim: The speaker suggests that, if a system ran on a GPU many thousands of times faster, it could access representations that make it fully transparent to itself and enable it to understand every mechanism producing its high-level experience and representations.

“So this is the inside perspective, right? And Buddhism, to me, is limited to this inside perspective because Buddhism is, uh, has been invented at a time where people were focusing very much on the experience of reality and how to adapt to this reality, but not how to split the atom, how to deal with the reality at the lowest level, at the foundations of physical reality.”

Claim: The speaker views Buddhism as focused on subjective experience and adapting to reality, rather than investigating the foundations of physical reality.

“So Buddhism is not so much interested in understanding physics. And in a way, this is more a Western tradition that starts with Aristotle, who thinks of the psyche as something that imposes itself on physics and interacts with it.”

Claim: The speaker contrasts Buddhism’s limited interest in physics with a Western tradition beginning with Aristotle, who viewed the psyche as acting on and interacting with physics.

“And to understand consciousness from the outside, we need to make a model of physics that can explain under which conditions systems can make arbitrary models, right?”

Claim: An external scientific account of consciousness requires a model of physics that explains the conditions under which systems can construct arbitrary models.

“Under which condition could a system build a model of what it's like to experience models at all? And this is something that computer science is doing today, that it looks at under representations, under which condition is a substrate able to hold a representation?”

Claim: Computer science studies the conditions under which a substrate can hold representations, including representations of the experience of modeling.

“Can hold a function that can be translated into behavior that can also be used to modify this function itself and the contents of that function. And this is, to me, super fascinating to look at.”

Claim: A substrate can hold a function that produces behavior and can itself be used to modify that function and its contents.

“Since I was a kid, I was fascinated by this concept of software, that you have these imaginary building blocks that are infinite in variety and shape, and there's no limitation to the, to these things that you can build and arrange except the memory of your computer that you can currently afford.”

Claim: Since childhood, the speaker has viewed software as infinitely varied and shapeable building blocks whose practical limit is the computer memory one can afford.

“And you can build a computer basically so large that it exceeds everything that you could ever want. And it's much larger than a human brain. And we are already there now that we have systems that are in principle larger than human brains and can hold more.”

Claim: The speaker says computers can be built with capacity exceeding human needs and that existing systems can in principle be larger than human brains and hold more.

“But we are still tinkering around and in the early days and have not found ways to evolve structures that build representation of the same complexity. And this outside perspective of consciousness is looking at what is consciousness doing in the brain?”

Claim: Although current systems can exceed the brain in capacity, they have not yet evolved structures that build representations of comparable complexity; an external account of consciousness investigates what consciousness does in the brain.

“Why does it emerge? What does it produce? And I suspect that this phenomenology of consciousness, that's what it's like, the internal causal structure of the reflection of what it would be like to take a certain perspective of something is that can already be built with today's systems.”

Claim: The speaker suspects that current systems can already construct the internal causal structure associated with the phenomenology of taking a particular perspective.

“By-- I'm agnostic with respect to whether the LLM is deepfaking it because it's been trained on human language that, uh, already contains a lot of these descriptions. Maybe we would need to build a system that learns language the old-fashioned way, that is developmental AI, that is learning in the same way as us.”

Claim: The speaker is uncertain whether an LLM’s apparent phenomenal experience is merely imitated from descriptions in its human-language training data.

“By-- I'm agnostic with respect to whether the LLM is deepfaking it because it's been trained on human language that, uh, already contains a lot of these descriptions. Maybe we would need to build a system that learns language the old-fashioned way, that is developmental AI, that is learning in the same way as us. And when it then spontaneously reports about its phenomenal experience, maybe that's much more convincing to us. But from the other perspective, from the outside perspective, I think that the LLM doesn't need our consciousness to function.”

Claim: A developmentally trained AI that learns language similarly to humans might be more convincing if it spontaneously reports phenomenal experience.

“And when it then spontaneously reports about its phenomenal experience, maybe that's much more convincing to us. But from the other perspective, from the outside perspective, I think that the LLM doesn't need our consciousness to function.”

Claim: The speaker thinks an LLM can function without consciousness.

“For you and me to talk, to write, to produce text, we need to be conscious. But the transformer doesn't need to be conscious to produce its output, and the, uh, n-neural network doesn't need to be conscious to produce this.”

Claim: The speaker says humans require consciousness to communicate and produce text, whereas transformers and neural networks can produce outputs without consciousness.

“Maybe consciousness and the way in which we have it is an artifact of running on a self-organizing system. So to me, the machine consciousness hypothesis has two parts.”

Claim: The speaker suggests that human-like consciousness may arise as an artifact of operating on a self-organizing system.

“One is the hypothesis about the role and nature and implementation of consciousness in organisms, right? It starts out with this description of what we agree on what consciousness is, and I hope that we can agree on it, on the phenomenology of consciousness.”

Claim: The speaker frames the machine-consciousness hypothesis as having two parts, the first addressing the role, nature, and implementation of consciousness in organisms and relying on a shared account of its phenomenology.

“And it's basically this experience of a now and perception of perception, and optionally the perspective of an observer that experiences itself and the act of observation, this first-person perspective.”

Claim: The speaker defines consciousness phenomenology as experiencing the present and perception of perception, optionally with a first-person observer experiencing itself and its own observation.

“There's also a minimal consciousness, which is simply attention on attention itself, where you don't have an observer present. And when we can agree on what we mean by the phenomenology of consciousness, we can then speculate about why consciousness is there and why we need to be conscious in order to function as human beings, as human minds, as intellects, and to develop.”

Claim: Agreement on the phenomenology of consciousness would allow speculation about why it exists and why it is needed for human functioning, intellect, and development.

“There's also a minimal consciousness, which is simply attention on attention itself, where you don't have an observer present. And when we can agree on what we mean by the phenomenology of consciousness, we can then speculate about why consciousness is there and why we need to be conscious in order to function as human beings, as human minds, as intellects, and to develop.”

Claim: Minimal consciousness consists of attention on attention itself without a present observer.

“And what we observe is that consciousness happens very early in our development. We are conscious as infants already, and if infants don't wake up, if this ignition of consciousness doesn't happen, if the lights don't go on, then we will not observe directed attention and learning and the formation of a world model in the infant and learning of language at some p- at some point.”

Claim: The speaker says human consciousness is present in infancy.

“And what we observe is that consciousness happens very early in our development. We are conscious as infants already, and if infants don't wake up, if this ignition of consciousness doesn't happen, if the lights don't go on, then we will not observe directed attention and learning and the formation of a world model in the infant and learning of language at some p- at some point.”

Claim: The speaker claims that without the onset of consciousness in an infant, directed attention, learning, world-model formation, and later language learning will not occur.

“Instead, the infant will remain vegetative. We will not turn into a human being. So consciousness seems to be crucial to train a brain like ours.”

Claim: The speaker says that without the onset of consciousness an infant would remain vegetative and not develop into a human being, suggesting consciousness is crucial to training a human-like brain.

“And it leads me to the hypothesis that consciousness might be a biological learning algorithm, a, a training algorithm for animal brains, at least animal brains.”

Claim: The speaker hypothesizes that consciousness may function as a biological learning algorithm that trains animal brains.

“I don't know if plants are conscious. That's something that we might have to find out in simulations once we understand what consciousness is. But if consciousness is basically a pattern in the communication between cells that leads naturally to the emergence of an observer that stabilizes itself as an observer to make learning happen and perception happen and so on, then we have a hypothesis about what consciousness is in the biological organism.”

Claim: The speaker proposes that if consciousness is a cell-communication pattern that produces a self-stabilizing observer, it could enable learning and perception and explain biological consciousness.

“I don't know if plants are conscious. That's something that we might have to find out in simulations once we understand what consciousness is. But if consciousness is basically a pattern in the communication between cells that leads naturally to the emergence of an observer that stabilizes itself as an observer to make learning happen and perception happen and so on, then we have a hypothesis about what consciousness is in the biological organism.”

Claim: The speaker is uncertain whether plants are conscious and suggests simulations may help investigate the question after consciousness is better understood.

“It's basically the simplest way to train a self-organizing system like us. And the way in which it trains it up automatically leads to this phenomenal experience, this experience of what it's like.”

Claim: The speaker proposes that consciousness may be a simple training mechanism for self-organizing systems whose operation automatically gives rise to phenomenal experience.

“This would be the hypothesis about biological consciousness, and this hypothesis might be wrong, right? Maybe it's something else altogether. Maybe it's something much weirder, but I don't see a reason to think that it's much more complicated than this.”

Claim: The speaker acknowledges that the biological-consciousness hypothesis may be wrong or that consciousness may be stranger than proposed, while seeing no reason to expect a much more complicated explanation.

“And if this hypothesis is true, then the second part, the machine consciousness hypothesis, might also be true, and this is that we might be able to reconstruct the conditions for self-organization in today's machines.”

Claim: If the proposed biological account of consciousness is correct, machines today might be able to recreate the conditions for self-organization and thereby support machine consciousness.

“So if we basically limit today's machines in such a way that they con- can only learn in self-organizing ways, and we give them tasks that are suitable in the large enough search space, then it might be possible that these machines discover consciousness in the same way as the brain of the infant discovers it.”

Claim: The speaker suggests that machines might discover consciousness in a way analogous to infant brains if they are limited to self-organizing learning and given suitable tasks in a sufficiently large search space.

“And we will see its presence by the fact that suddenly the system undergoes some phase transition that enables it to learn. And at some point, it will train itself up into something that will also naturally report on phenomenology and so on.”

Claim: The speaker predicts that machine consciousness, if it emerges, would be indicated by a learning-enabling phase transition and later by self-training and spontaneous reports of phenomenology.

“So in this way, we won't have a Turing test for consciousness because the Turing test is about performance, where Turing's test does not care about how you produce the dialogue.”

Claim: The speaker argues that a Turing test cannot establish consciousness because it assesses performance rather than the process producing dialogue.

“For Turing, an LLM is sufficient as long as the LLM is good enough at writing a poem and, uh, revising it, and at solving mathematical problems and playing chess, and so on and so on.”

Claim: The speaker says that under Turing’s performance-oriented standard, an LLM could suffice if it performed well at writing and revising poems, mathematics, chess, and similar tasks.

“Turing doesn't care about whether the, um, machine is also conscious, whether it has the same internal mechanisms. But for consciousness, we do care whether there is an actual experience of what it's like in the system.”

Claim: The speaker distinguishes performance testing from consciousness: Turing’s test ignores a machine’s consciousness and internal mechanisms, while consciousness depends on whether the system has subjective experience.

“And so this is more an not a test, but it's an interpretation. And the interpretation here is we interpret the biological consciousness as something that forms under these and these conditions for that and that purpose.”

Claim: The speaker characterizes the approach as interpreting biological consciousness in terms of the conditions under which it forms and the purposes it serves, rather than testing it directly.

“And if that is correct, then we might be able to recreate this into a causal system that has an equivalent structure. And these two hypotheses, the machine consciousness hypothesis and the biological consciousness hypothesis, can be wrong independently of each other, of course, right?”

Claim: If the biological account of consciousness is correct, an equivalent structure might be recreated in an artificial causal system, but the biological and machine-consciousness hypotheses can fail independently.

“It might be that my idea of what biological consciousness is, is correct, but we're still not able to recreate it in the artificial substrate. And this could have multiple reasons.”

Claim: The speaker allows that the biological theory of consciousness may be correct while its recreation in an artificial substrate remains impossible.

“For instance, maybe our computers are not fast enough. Maybe we require too many of the moving parts of the individual cells to make consciousness happen.”

Claim: The speaker suggests that artificial consciousness might be unattainable because computers are too slow or because reproducing it requires too many cellular components.

“Maybe, um, just having a pattern across neurons and so on is not good enough. But why should I expect this to happen?”

Claim: The speaker allows that a pattern distributed across neurons may not by itself be sufficient for consciousness.

“Because the current, um, AIs are fast enough to produce, for instance, game engines that look very much like our own internal experience, and they can produce output that happens at roughly the same rate as human brains can make it.”

Claim: The speaker says current AI systems can produce experience-like game engines and outputs at approximately the rate of human brains.

“And they can also process inputs at roughly at the rate at which human brains can produce input, give or take. So we are now with our computers in this ballpark and with our algorithms in the ballpark of what human brains can be doing.”

Claim: The speaker says current computers and algorithms operate in the approximate range of human brains for processing input and producing output.

“And so is this dream within a dream, this dream of actually dreaming a dream much, much more important-- uh, difficult to achieve? Does it require much more compute than producing the contents of the dream? And I doubt that this is the case. If it's not the case, that it's more complicated to dream of dreaming than just to have the contents of the dream that the LLM can already produce, right, or the frontier model, then, um, we should expect that comp-- today's computers are fast enough.”

Claim: The speaker doubts that generating the experience of dreaming requires substantially more computation than generating dream contents.

“And I doubt that this is the case. If it's not the case, that it's more complicated to dream of dreaming than just to have the contents of the dream that the LLM can already produce, right, or the frontier model, then, um, we should expect that comp-- today's computers are fast enough.”

Claim: If representing the experience of dreaming is no more computationally complex than producing dream contents, the speaker expects today’s computers to be fast enough to do it.

“Another way in which the project can fail is that the search space is too large, that the main ways in which a biological system can train itself, there are too many possibilities to discover them indepen-in-uh independently.”

Claim: The speaker identifies an excessively large search space, with too many possible self-training methods, as a possible obstacle to machines independently discovering consciousness.

“And maybe we don't know how long nature took to discover it. Maybe consciousness is indeed discovered in every individual brain. But maybe it's not. Maybe the, um, switch from single-cellular life to multicellular life took so long because morphogenesis, this algorithm that allows organisms to become coherent at scale, is so difficult to discover.”

Claim: The speaker suggests that the transition to multicellular life may have taken a long time because the morphogenesis algorithm that creates organism-level coherence is difficult to discover.

“And maybe we don't know how long nature took to discover it. Maybe consciousness is indeed discovered in every individual brain. But maybe it's not. Maybe the, um, switch from single-cellular life to multicellular life took so long because morphogenesis, this algorithm that allows organisms to become coherent at scale, is so difficult to discover.”

Claim: The speaker does not know how long nature took to discover consciousness or whether each individual brain discovers it independently.

“And Mike Levin has this hypothesis that this algorithm for morphogenesis, the algorithm that makes an organism coherent with itself, is exactly the same one as the one that happens in the mind when it becomes coherent with itself.”

Claim: Mike Levin’s hypothesis is that organism-level coherence and mental coherence arise from the same algorithm.

“So this coherence in, uh, maximizing pattern that we observe in our own mind to play out when we are conscious is maybe the same pattern at a different timescale that happens in the organism.”

Claim: The speaker suggests that the pattern producing coherence in conscious minds may also produce organism-level coherence on a different timescale.

“And so consciousness and life, multicellular life at least, could be the same. And if that is the case, maybe it took a very, very long time to discover and you don't discover it independently in your consciousness, but you get infected by your mother in utero or you basically vibe with your biological environment, with all the biological organisms around you, and they basically nudge your mind to discover consciousness, so it's much, much easier.”

Claim: The speaker speculates that consciousness and multicellular life may share a process, in which case consciousness may be difficult to discover independently and may be facilitated by prenatal and broader biological environments.

“And without this nudge, maybe the search space is too large for our computers to become conscious, and maybe we would need to surround our GPUs with lots of antennas that can pick up the vibe from the biological environment in the right way, so it gets inspired by it and discovers the same thing.”

Claim: The speaker speculates that computers might need cues from the biological environment to overcome a search space too large for them to discover consciousness unaided.

“It's a very poetic idea. But again, why would this be the case? It's not necessarily the case, right? It would not be my natural expectation that this is the way.”

Claim: The speaker emphasizes that the proposed need for biological cues is not established and is not the speaker’s default expectation.

“So my null hypothesis should be consciousness is easy enough to be discovered in the individual brain, and consciousness is an algorithm that is not much more complicated than, say, the transformer.”

Claim: The speaker’s default hypothesis is that consciousness can be discovered within an individual brain and is not much more complicated than a transformer.

“And the transformer is a beautiful example. The transformer is a thing that is so easy to understand that you don't need to study more than a week if you know computer science to get it, how it works, right?”

Claim: The speaker says that someone who knows computer science can understand how a transformer works with roughly a week or less of study.

“It's, it's quite beautiful. And the stuff the transformer is learning is much more complicated than a transformer itself. All the common sense knowledge that you extract from text, all the algorithms, all the procedures and so on, they are so intricate.”

Claim: The speaker says that the common-sense knowledge, algorithms, and procedures learned from text are more intricate than the transformer that learns them.

“There are so many things that this simple algorithm can learn, and there is no obvious limit to what it can learn. And so maybe consciousness is like this.”

Claim: The speaker says a relatively simple transformer can learn a wide range of complex material with no obvious limit, and suggests consciousness may have a similar relationship to its underlying algorithm.

“It's an algorithm that is not trying to, um, learn how to predict the next token. Maybe it's an algorithm that's trying to make working memory coherent with itself.”

Claim: The speaker proposes that consciousness may operate by making working memory internally coherent, rather than by next-token prediction.

“It's some kind of consensus algorithm across mental representations. So maybe it's a very different beast. Maybe it's something that converges much faster.”

Claim: The speaker suggests that consciousness may function as a consensus algorithm across mental representations and may converge relatively quickly.

“Can also turn out that it's very efficient to run-- inefficient to run it on GPUs, and it's only something that runs efficiently on brains. That's something that's really open to research and that I would be very excited if we can see it at some point.”

Claim: The speaker says it remains an open question whether consciousness is inefficient on GPUs and efficient only on biological brains.

“That's a super interesting question. I suspect that at some point it might be cheaper and easier to start again because true machine perception has never been tried.”

Claim: The speaker suspects that starting again from scratch might eventually be cheaper and easier because true machine perception has never been tried.

“The data that we give the LLM is stuff that is either produced by humans or it's made for human media consumption, which means it's adapted for our mostly visual and auditory sensors.”

Claim: The speaker says that LLM training data is either human-produced or designed for human media consumption, and is therefore adapted to humans’ mostly visual and auditory sensors.

“But there is no reason why you should be limited to the human sensorium, right? You could have something that has much higher ram-- frame rates and looking at very different frequency ranges all at once.”

Claim: The speaker argues that an AI system need not be limited to human sensory capabilities and could use much higher frame rates and observe different frequency ranges simultaneously.

“And so you can, in principle, build a much richer, more detailed world model that is coarse graining in a different way. Um, the-- You mentioned the universality hypothesis indirectly, right?”

Claim: The speaker says that systems with broader sensory input could, in principle, build a richer and more detailed world model using a different form of coarse-graining.

“This work by Chris Olah. What, uh, his team discovered is that when you are looking at the different vision models, you find that they all learn the same features.”

Claim: The speaker reports that Chris Olah’s team found that different vision models learn the same features.

“And it's regardless of the architecture and the particular learning algorithm and even the details of the data. And so the idea behind this is that the structure of your model does not really depend on the algorithm or on the data, but it, um, really depends on the structure of the problem that you're trying to learn.”

Claim: The speaker says that learned features are similar across vision models despite differences in architecture, learning algorithm, and data details, suggesting that model structure depends on the problem’s structure rather than those choices.

“And you find that this-- despite the learning and the transformer algorithm or state space model or other perceptron algorithms, uh, is very, very different from the way in which the brain is learning.”

Claim: The speaker says that transformer, state-space, and other perceptron learning algorithms operate very differently from the brain’s learning process.

“The visual architecture of the brain is also very similar to a computer vision model. That is, you have at the lowest level, um, colors and, uh, spatial frequencies, and then you build this into textures that are combined, and then at some point you have, uh, geometries and so on.”

Claim: The speaker says that the brain’s visual architecture resembles computer vision models, progressing from colors and spatial frequencies to textures and then geometries.

“And the deep question is- At which level does this stop? If you, the universality hypothesis would be universally true, right? It would mean that if you train a model on your own in input and output, and you train it for long enough, you will wake up inside of that model. And it's a quite mind-blowing idea to me, and I don't know if that is actually the case, and there might be a reason why it's not the case.”

Claim: The speaker considers it uncertain whether universal convergence would mean that a model trained long enough on its own inputs and outputs would reproduce a mind; he finds the possibility mind-blowing and says it may not hold.

“Maybe language is insufficiently constraining the causal structure of the system that is producing language. Maybe there are many ways in which the same set of linguistic utterances that a human being is producing can be produced.”

Claim: The speaker suggests that language may insufficiently constrain the causal structure of its producing system, allowing many different systems to generate the same human utterances.

“But if you think about it, a human being ultimately is just an end-to-end trained learning system, end-to-end trained over an evolutionary time span, but also your biography.”

Claim: The speaker characterizes a human being as an end-to-end trained learning system shaped by both evolution and individual biography.

“And these together produce your behavior, and you would assume that biology roughly gets to the simple solution to produce your output given your input, right?”

Claim: The speaker says that evolution and personal biography together shape human behavior, and assumes biology roughly finds a simple way to produce behavior from input.

“And so if you take a neural network that is trained to produce similar output when given the same input, you would also think that this new algorithm that we have implemented is not doing any kind of magic that ends up with something that is way more complicated to produce the same output.”

Claim: The speaker expects a neural network trained to match human input-output behavior not to rely on a process that is inexplicably much more complicated than the process producing the same behavior.

“But ultimately, it's going to converge to a structure that is similar to yours, to a similar causal structure. But there might be conditions when that is not the case. For instance, maybe certain things run well on a brain, but they don't run well on a GPU or vice versa.”

Claim: The speaker expects a system trained to match human input-output behavior to converge toward a similar causal structure, but says this may fail when a computation runs well on a brain but poorly on a GPU, or the reverse.

“There is a lot of stuff that is easy to run on a GPU and almost impossible to run on a brain, so the brain is going to use workarounds. In fact, I believe that our GPUs are more powerful than brains.”

Claim: The speaker says that many computations are easy on GPUs but nearly impossible for brains, leading brains to use workarounds, and believes GPUs are more powerful than brains.

“Basically, the set of algorithms that can run on GPUs is not limited in any way. You can impose the same constraints as you have in biology in principle. These aren't the limits of the resolution, the speed, the size of the memory that you have, and so on.”

Claim: The speaker says GPU algorithms are not inherently restricted to biological constraints, although those constraints—such as resolution, speed, and memory limits—can be imposed in principle.

“But if you have a hardware that is fast enough, you can simulate an arbitrary natural system to an arbitrary degree of fidelity, whereas your brain cannot simulate a digital computer simply because it's very mushy.”

Claim: The speaker says sufficiently fast hardware can simulate any natural system to any desired degree of fidelity, while a brain cannot simulate a digital computer because the brain is too imprecise or “mushy.”

“So your brain is limited to the way in which it works. And so if it turns out that our own consciousness is an artifact of our brain trying to make complicated intelligent stuff work, and, uh, the computer does not need the same artifact, it could be that the computer comes up with something that is more, um, clever and more elegant than our own consciousness.”

Claim: The speaker suggests that if consciousness is an artifact of the brain’s effort to support complex intelligence, a computer might not need consciousness and could develop a more clever and elegant alternative.

“... wrong about Picbreeder, but the way I understand that it works, it's using a neural network that has been specifically tinkered for making representations like this.”

Claim: Picbreeder uses a neural network specifically modified to produce the kinds of representations being discussed.

“So it's something that is not simply using ReLU for its output, but it might be using Gaussians or, um, sine f- uh, functions and so on as the output function.”

Claim: Picbreeder's neural network may use Gaussian or sine output functions instead of simply using ReLU as its output function.

“They can be combined to make patterns that repeat, or they can be make, uh, a pattern that are symmetric. And so basically, they tweak the neural network that they're using from the beginning to be suitable for this.”

Claim: Picbreeder's neural network is designed from the outset so its functions can be combined to produce repeating or symmetric patterns.

“The other one is that the way in which Picbreeder is trained is not automatically. It's trained with a population of human raters. So the fitness function is basically humans who look at the output of Picbreeder and say, "This is a good start, and this is how we should be proceeding."”

Claim: Picbreeder's training uses human raters to evaluate outputs and guide subsequent development rather than proceeding automatically.

“And so I think that this process of the construction of the pictures together with the mechanism that produces the patterns is biasing Picbreeder to make stuff that is more organic, that is more controllable, and so on.”

Claim: The combination of Picbreeder's picture-construction process and pattern-producing mechanism biases it toward more organic and controllable outputs.

“But I believe this could be imitated, right? We could automate the procedure by the, uh, which human beings are selecting pictures by using a diffusion model that is rating the output for similarity to human output.”

Claim: Picbreeder's human picture-selection procedure could be automated with a diffusion model that rates outputs by their similarity to human output.

“So we, we could use an AI that is in principle already tweaked to imitate the behavior of the human juries that are performing this thing.”

Claim: An AI already tuned to imitate the behavior of Picbreeder's human juries could be used to automate their selection procedure.

“And we could also evolve architectures that have a larger variety of activation functions that are more suitable for this and ultimately outperform Picbreeder by this.”

Claim: Evolving architectures with a wider variety of activation functions better suited to the task could ultimately outperform Picbreeder.

“I think that Sutton's bitter lesson is also going to be true for this one. So if we basically take the human out of the equation and instead just evolve the search space or search the, uh, the space of solutions automatically, we are also going to end up with something that's more elegant than the hand-tinkered solution in the case of Picbreeder.”

Claim: The speaker expects Sutton's bitter lesson to apply to Picbreeder: automatically searching or evolving the solution space without human involvement will yield a more elegant result than its hand-tinkered solution.

“In the case of Picbreeder, we have a system that is inherently limited because these networks are very small, right? They're very sparse. And so you have relatively few parameters that you need to optimize, and you set up the model in such a way that most of the parameters that you're playing with are meaningful.”

Claim: Picbreeder uses small, sparse networks with relatively few parameters to optimize, and most of those parameters are meaningful.

“On the other hand, it's limited because you're not going to get Picbreeder to produce something that looks arbitrarily complicated. And if you want to move in these realms, then you have to have tools that are much more complex.”

Claim: Picbreeder cannot produce arbitrarily complicated outputs, and working in more complex domains requires more complex tools.

“It's similar to, um, trying to produce very good user-generated content using Blender versus Minecraft. And so with Minecraft, it's very easy for people to produce meaningful stuff, and it's almost difficult not to produce something with Minecraft that looks somewhat meaningful if-- and that you can play around with and walk around in and have fun in.”

Claim: Compared with Blender, Minecraft makes it easier for people to create meaningful user-generated content.

“And if you have a complex 3D program, the g- uh, inverse is true because most of the stuff that you're doing with this complicated thing is not going to be interesting, and it's not going to look e-exciting.”

Claim: Most outputs made with a complex 3D program will not be interesting or exciting.

“But it will look like, "Oh my god, s- uh, this person has no ability to use its tools." And so, uh, th-that is, um, one of the issues when we are building learning systems. If we make them very simple, it's easier to control the complexity, similar as it is in art. And it can be exciting to work with constraints, and if you remove the constraints, the task becomes much more frustrating because you would have to suddenly, uh, to conquer an enormously large space, and the space might be simply too large for human beings.”

Claim: Using very simple tools in learning systems makes their complexity easier to control.

“If we make them very simple, it's easier to control the complexity, similar as it is in art. And it can be exciting to work with constraints, and if you remove the constraints, the task becomes much more frustrating because you would have to suddenly, uh, to conquer an enormously large space, and the space might be simply too large for human beings.”

Claim: Working within constraints can be exciting.

“If we make them very simple, it's easier to control the complexity, similar as it is in art. And it can be exciting to work with constraints, and if you remove the constraints, the task becomes much more frustrating because you would have to suddenly, uh, to conquer an enormously large space, and the space might be simply too large for human beings.”

Claim: Removing constraints can make a task much more frustrating because the resulting solution space may be too large for humans to navigate.

“I mean, assuming you're right about the convergent representations in, in language models, I guess it's not a foregone conclusion that the consciousness program would be convergent in the same way. It might be divergent. And how far down the stack does consciousness go?”

Claim: The consciousness program is not guaranteed to converge in the same way as representations in language models and might instead be divergent.

“I think this, uh, is an experimental question, right? Uh, all cultures outside of ours are animist cultures, and animism is the belief that the difference between living and dead systems is that the living systems are possessed by spirits, and spirits are intrinsically agentic and capable of experience.”

Claim: How consciousness works is an experimental question.

“I think this, uh, is an experimental question, right? Uh, all cultures outside of ours are animist cultures, and animism is the belief that the difference between living and dead systems is that the living systems are possessed by spirits, and spirits are intrinsically agentic and capable of experience.”

Claim: Animism holds that living systems differ from dead systems because living systems are possessed by spirits that are intrinsically agentic and capable of experience.

“I think this, uh, is an experimental question, right? Uh, all cultures outside of ours are animist cultures, and animism is the belief that the difference between living and dead systems is that the living systems are possessed by spirits, and spirits are intrinsically agentic and capable of experience.”

Claim: All cultures outside the speaker's culture are animist cultures.

“And the scientific worldview rejects the notion of spirits as unscientific, right? Despite the fact that there's clearly some spirit running on you and me that is agentic and capable of experience.”

Claim: The scientific worldview rejects the notion of spirits as unscientific, although the speaker asserts that humans have an agentic spirit capable of experience.

“So there is an existence proof for such spirits, right? The question is, how does the spirit work? The claim that the spirit is immortal, that, uh, it's, uh, cannot be explained using physics and so on, is not necessarily a claim that is intrinsic to animism.”

Claim: The speaker treats humans' agentic, experience-capable spirit as an existence proof for spirits of that kind.

“So there is an existence proof for such spirits, right? The question is, how does the spirit work? The claim that the spirit is immortal, that, uh, it's, uh, cannot be explained using physics and so on, is not necessarily a claim that is intrinsic to animism.”

Claim: The claim that spirits are immortal or inexplicable through physics is not necessarily intrinsic to animism.

“The spirit doesn't have to be supernatural. Instead, I think that it is natural. And I suspect that Anil Seth's position of his biological naturalism by making a claim that biological systems are intrinsically different from any other system that is not biological is actually a supernaturalism, right?”

Claim: The speaker suspects that Anil Seth's biological naturalism amounts to supernaturalism because it treats biological systems as intrinsically different from nonbiological systems.

“The spirit doesn't have to be supernatural. Instead, I think that it is natural. And I suspect that Anil Seth's position of his biological naturalism by making a claim that biological systems are intrinsically different from any other system that is not biological is actually a supernaturalism, right?”

Claim: The speaker believes that the spirit is natural rather than necessarily supernatural.

“He is a biological supernaturalist. He believes that biological systems cannot be fully naturalized, which means you cannot translate them into mathematical models.”

Claim: The speaker says Anil Seth believes biological systems cannot be fully naturalized or translated into mathematical models, and characterizes this position as biological supernaturalism.

“Whereas the dead stuff, you can translate them into mathematical models. And so I would say that, uh, physics, which is the attempt to translate everything into mathematical models, right, is, uh, not stopping at biological systems.”

Claim: The speaker describes physics as an effort to translate everything, including biological systems, into mathematical models.

“Whereas the dead stuff, you can translate them into mathematical models. And so I would say that, uh, physics, which is the attempt to translate everything into mathematical models, right, is, uh, not stopping at biological systems.”

Claim: The speaker says dead matter can be translated into mathematical models.

“Because the difference with these biological systems is that they're possessed by software, and the software is self-organizing. And in order to be self-organizing, it's possible that it needs to be capable of experience, that it needs to monitor itself across multiple dimensions and observe itself in the act of monitoring to make that self-organization happen.”

Claim: The speaker suggests that self-organizing software may require experience, including multidimensional self-monitoring and observing its own monitoring, to self-organize.

“Because the difference with these biological systems is that they're possessed by software, and the software is self-organizing. And in order to be self-organizing, it's possible that it needs to be capable of experience, that it needs to monitor itself across multiple dimensions and observe itself in the act of monitoring to make that self-organization happen.”

Claim: The speaker says biological systems differ from other systems because they are possessed by self-organizing software.

“And so maybe this pattern has to be discovered in nature to make life possible, at least multicellular life with a certain complexity. It seems to be definitely the case for animal brains.”

Claim: The speaker suggests that a self-organizing, experience-capable pattern may need to occur in nature for life to be possible, at least for multicellular life above a certain complexity.

“And so maybe this pattern has to be discovered in nature to make life possible, at least multicellular life with a certain complexity. It seems to be definitely the case for animal brains.”

Claim: The speaker says the pattern is definitely present in animal brains.

“For the others, we don't know. For the, uh, for, uh, we don't know how complex the minds of plants are. We know that plants are able to communicate within themselves and across each other to some degree, but we don't know how smart they are, and we don't know whether they are sentient in the sense that they experience themself in the act of experience.”

Claim: The speaker says the complexity of plant minds is unknown.

“For the others, we don't know. For the, uh, for, uh, we don't know how complex the minds of plants are. We know that plants are able to communicate within themselves and across each other to some degree, but we don't know how smart they are, and we don't know whether they are sentient in the sense that they experience themself in the act of experience.”

Claim: The speaker says it is unknown how intelligent plants are and whether they are sentient in the sense of experiencing themselves experiencing.

“For the others, we don't know. For the, uh, for, uh, we don't know how complex the minds of plants are. We know that plants are able to communicate within themselves and across each other to some degree, but we don't know how smart they are, and we don't know whether they are sentient in the sense that they experience themself in the act of experience.”

Claim: The speaker says plants can communicate internally and with other plants to some degree.

“Aristotle saw that the plants are relatively dumb. He basically said there are three levels of souls in nature, and he basically said that plants have the lowest one. There's this, this, uh, level that is vegetative, that is capable of nurturing and of growth, but it does not experience the world around us.”

Claim: According to the speaker's account of Aristotle, plants have the lowest of three kinds of soul: a vegetative soul that supports nurturing and growth but does not experience the world around it.

“It doesn't sense. And so the sensate mind, the sensate soul that is, uh, also capable of making decisions and having preferences, that is something that only animals have.”

Claim: According to the speaker's account of Aristotle, only animals have a sensate soul that can sense, make decisions, and have preferences.

“It doesn't sense. And so the sensate mind, the sensate soul that is, uh, also capable of making decisions and having preferences, that is something that only animals have.”

Claim: According to the speaker's account of Aristotle, plants do not have sensation.

“And human beings are animals that in addition have a rational soul, a rational soul that is able to reflect and make sense of itself and the world in which, in a way in which animals cannot.”

Claim: According to the speaker's account of Aristotle, humans are animals with an additional rational soul that can reflect and understand itself and the world in a way other animals cannot.

“And I think that's very much aligned with today's scientific worldview, which is not surprising because our today's scientific worldview is very much, uh, invented by Aristotle and goes back to him and is an extension of his ideas and a systematization and exploration of his ideas.”

Claim: The speaker believes today's scientific worldview aligns strongly with Aristotle's ideas because it was largely developed from and extends, systematizes, and explores those ideas.

“But we don't know how smart plants are, right? Maybe Mike Levin is correct in that every organism is in some sense sentient, and sentience is something that is universal to the, uh, the spirits, the self-organizing software nature.”

Claim: The speaker says the intelligence of plants is unknown.

“But we don't know how smart plants are, right? Maybe Mike Levin is correct in that every organism is in some sense sentient, and sentience is something that is universal to the, uh, the spirits, the self-organizing software nature.”

Claim: The speaker allows that Mike Levin may be right that every organism is sentient in some sense and that sentience is universal to nature's spirits or self-organizing software.

“And I'm agnostic with respect to that. I have no idea whether that's the case, right? I, I don't know. I don't have evidence to the contrary. I also don't have strong evidence in favor, and I think this is still an open question.”

Claim: The speaker is agnostic about whether every organism is sentient: he lacks strong evidence for or against the idea and considers it an open question.

“And to me, it's exciting to explore this question and surprising that so few people are actually dedicating their life to answering it.”

Claim: The speaker finds the question of whether organisms are sentient exciting to investigate and is surprised that few people devote their lives to answering it.

“The issue with Searle is that there is very little for me to engage. There is a claim, but no argument. And a similar thing is happening with, uh, Anil's point that he has a claim, but there is no argument that I can engage with, right?”

Claim: Joscha Bach says Searle makes a claim about consciousness without providing an argument for Bach to engage with.

“The issue with Searle is that there is very little for me to engage. There is a claim, but no argument. And a similar thing is happening with, uh, Anil's point that he has a claim, but there is no argument that I can engage with, right?”

Claim: Joscha Bach says Anil Seth makes a claim without providing an argument for Bach to engage with.

“There is nothing, no experiment that could convince him to the contrary, because he's ultimately only working with a certain intuition that he tries to explore more.”

Claim: Joscha Bach says no experiment could convince John Searle otherwise because Searle is ultimately working from an intuition that he seeks to explore.

“And so the way in which I can deal with, uh, the arguments of Searle and Anil Seth is ultimately only a psychological one.”

Claim: Joscha Bach says he can address John Searle's and Anil Seth's arguments only by examining them psychologically.

“I would have to look at why does this seem convincing to them, and why doesn't it seem to be convin-- not convincing to me, right? So we look at the structure of our own minds, and ultimately it comes down to a difference in epistemology.”

Claim: Joscha Bach says the difference in how John Searle, Anil Seth, and Bach find arguments convincing ultimately comes down to differences in epistemology.

“And what I observe is that a lot of philosophers, the atoms of their language when they describe themselves, th-their own hermeneutics, ends at agents.”

Claim: Joscha Bach observes that many philosophers treat agents as the endpoint of the basic concepts they use to describe themselves and their hermeneutics.

“It ends at the experience of something. And, uh, similar to Buddhism, where Buddhism is not, uh, disassembling the world into automata as computer scientists are doing.”

Claim: Joscha Bach says an introspective philosophical approach ends with experience and contrasts it with Buddhism, which does not break the world down into automata in the way computer scientists do.

“Computer scientists basically resolve everything into, um, states that can be described by their transitions. And that's a very fundamental way of thinking about reality, and it's a tradition that finds itself throughout, um, the philosophy as this attempt to mathematize the world, of describing minds as dynamical mathematical objects.”

Claim: Joscha Bach says computer scientists model reality as states characterized by their transitions.

“Computer scientists basically resolve everything into, um, states that can be described by their transitions. And that's a very fundamental way of thinking about reality, and it's a tradition that finds itself throughout, um, the philosophy as this attempt to mathematize the world, of describing minds as dynamical mathematical objects.”

Claim: Joscha Bach says the philosophical project of mathematizing the world includes modeling minds as dynamical mathematical objects.

“And, um, this finds itself in its modern incarnation in computer science, where mathematics is being described from the ground up by the tools that we can make to build arbitrary arithmetic and geometry from simple operations over bit patterns.”

Claim: Joscha Bach says modern computer science constructs arithmetic and geometry from simple operations on bit patterns, describing mathematics from the ground up.

“And so our world is constructed from the ground up from smallest elements that themselves are not agents yet, that are not minds yet, that are not experiences yet.”

Claim: Joscha Bach says the world is built from basic elements that are not themselves agents, minds, or experiences.

“And there is another tradition in philosophy that is basically working more from this introspection, and this-- only the stuff that I can introspect over seems to be valid, and the stuff that computer scientists are doing in their basement is unrelated to it.”

Claim: Joscha Bach says an introspection-centered philosophical tradition treats only what can be introspected as valid and regards computer scientists' work as unrelated.

“And Searle and Anil Seth, they both don't see the way in which to connect automata to things that can feel, that can reason, that can make models of reality.”

Claim: Joscha Bach says John Searle and Anil Seth do not see a path from automata to entities capable of feeling, reasoning, and modeling reality.

“And computer scientists see such a path. I think that's the main difference. And, uh, we-- But we can create evolution in computers, right?”

Claim: Joscha Bach says computer scientists see a path from automata to entities capable of feeling, reasoning, and modeling reality.

“And computer scientists see such a path. I think that's the main difference. And, uh, we-- But we can create evolution in computers, right?”

Claim: Joscha Bach says it is possible to create evolution in computers.

“We can create the conditions for software that is not written by us, and that's already true for the models that we are building with transformers.”

Claim: Joscha Bach says people can create conditions for software not written by humans, and that transformer models already exemplify this.

“They're not written by human beings. They're statistical aggregates that are woven, that are grown. And in many ways, the computer models that we're playing with today, they're, they're a result of growth, and they're pruned, and they are gardened and, uh, chaperoned and, uh, trained in, in a way that is much more in which we, uh, observe this happening in biology than it happened when we were writing computer programs by hand.”

Claim: Joscha Bach characterizes transformer models as statistical aggregates produced through growth, pruning, gardening, chaperoning, and training rather than being written by humans.

“They're not written by human beings. They're statistical aggregates that are woven, that are grown. And in many ways, the computer models that we're playing with today, they're, they're a result of growth, and they're pruned, and they are gardened and, uh, chaperoned and, uh, trained in, in a way that is much more in which we, uh, observe this happening in biology than it happened when we were writing computer programs by hand.”

Claim: Joscha Bach says today's computer models are developed in a way more like biological growth than like hand-written computer programming.

“That's also an interesting observation about human minds. During the pandemic, we observed that a lot of people, when they were starting to work from home, became dysfunctional.”

Claim: Joscha Bach says many people became dysfunctional after they started working from home during the pandemic.

“And this seems to be that many people can only be stabilized due to their environment, that they do require an environment that is constantly affirming their reality.”

Claim: Joscha Bach says many people depend on an environment that continually affirms their reality in order to remain stable.

“And if you take that away, they can go off the rails. And if you have an LLM that is interacting with them, then there can be an interaction between the LLM and the human being that pulls the human being off the rails if they don't have a strong foundation in their own reality.”

Claim: Joscha Bach says people who depend on environmental affirmation can go off the rails when that support is removed.

“And if you take that away, they can go off the rails. And if you have an LLM that is interacting with them, then there can be an interaction between the LLM and the human being that pulls the human being off the rails if they don't have a strong foundation in their own reality.”

Claim: Joscha Bach says an LLM-human interaction can destabilize a person who lacks a strong foundation in their own reality.

“And, uh, I think that the LLMs have now reached a point where they're m-memetically more powerful than many human mi-minds are.”

Claim: Joscha Bach thinks LLMs are now memetically more powerful than many human minds.

“There are things that the LLM cannot do at this point, that where they're still hallucinating, where they're still in- more incoherent than experts in a given field, but they are much better at prompt completion than human beings are.”

Claim: Joscha Bach says LLMs still hallucinate and remain more incoherent than experts in a given field.

“There are things that the LLM cannot do at this point, that where they're still hallucinating, where they're still in- more incoherent than experts in a given field, but they are much better at prompt completion than human beings are.”

Claim: Joscha Bach says LLMs outperform human beings substantially at prompt completion.

“And many human beings are professional prompt completers, right? They're incapable of having a single original creative thought. Instead, they are trying to serve expectations of the environment.”

Claim: Joscha Bach says many people function as prompt completers, unable to produce original creative thoughts and instead responding to their environment's expectations.

“They are picking up on the statistical patterns in the language of the environment, and then you tell them a headline, and they write an article. And, uh, now you give them an LLM.”

Claim: Joscha Bach says people who act as prompt completers absorb statistical patterns in their environment's language and can turn a headline into an article.

“How is the environment going to change, right? Once they talk more to LLMs than they talk to other human beings because the LLM is such a useful tool for them to do their prompt completion, what's going to happen to them?”

Claim: Joscha Bach asks what will happen to people who interact more with LLMs than with other humans because LLMs are useful for their prompt-completion behavior.

“It's a very interesting thing what happens to us weak-minded human beings when we are hypnotized by things that are more hypnotic than our mind is to itself.”

Claim: Joscha Bach says it is an important question what happens to people he calls weak-minded when they are influenced by things more hypnotic than their own minds.

“I think physicalism is a certain hypothesis, and it comes down to there being a causally closed mechanical layer to reality that can be researched using experiments in physics labs, and, uh, that this causally closed mechanical layer of reality is not moved by our intentions, right?”

Claim: Physicalism holds that reality has a causally closed mechanical layer, investigable through physics experiments, that is not affected by human intentions.

“It does not care about what you want. And the second part of physicalism is that we are directly supervening over this mechanical, causally closed world.”

Claim: Physicalism holds that minds or human beings directly supervene on the mechanical, causally closed world.

“So we are-- there's not some parent universe that contains a server where a nerd is running a simulated universe that is some kind of Minecraft, and we live inside of this Minecraft. So, uh, it's, it's not some simulation hypothesis. And so physicalism is these two components. But to me, physics is not something that has revealed itself directly.”

Claim: Physicalism is distinct from the hypothesis that our universe is a simulation run by an external programmer.

“So, uh, it's, it's not some simulation hypothesis. And so physicalism is these two components. But to me, physics is not something that has revealed itself directly.”

Claim: Physics is not something that has been directly revealed to human observers.

“So nobody has ever seen matter and energy directly. Instead, matter and energy are attempts to systematize observations, and our observations ultimately all come down to discernible differences.”

Claim: Matter and energy are conceptual systems for organizing observations, which ultimately consist of discernible differences.

“So nobody has ever seen matter and energy directly. Instead, matter and energy are attempts to systematize observations, and our observations ultimately all come down to discernible differences.”

Claim: Human observers have never directly seen matter or energy.

“And the technical term for discernible differences is information. So to me, information is fundamental when we talk about reality. Reality is about differences in information, and physics is a certain layer of regularity in the differences in information that we have discovered.”

Claim: Physics describes a layer of regularities discovered in differences in information.

“And the technical term for discernible differences is information. So to me, information is fundamental when we talk about reality. Reality is about differences in information, and physics is a certain layer of regularity in the differences in information that we have discovered.”

Claim: Information, understood as discernible differences, is fundamental to reality, which consists of differences in information.

“So to me, matter is not at the bottom. The bottom is really all these, uh, differences that we observed and which we experimentally and practically empirically found to be so regular that we-- that they can be translated into this mechanical mo- world of physics.”

Claim: Matter is not fundamental; the foundation is observed differences whose empirical regularity allows them to be represented in the mechanical world described by physics.

“And then on top of it, we find at some point the organization into organisms, and the organisms are producing minds, and the minds are simulating us and simulating our sense of realness, of our sense of being in a world with colors and sounds and people and so on.”

Claim: Organisms emerge from underlying physical organization and produce minds that simulate the experienced world, including colors, sounds, and people.

“And in this world, intentions matter. The psychological reality has things like will and emotion and despair and desire and, uh, all these things that, um, make us tick.”

Claim: In psychological reality, intentions matter, and experiences include will, emotion, despair, and desire.

“And, uh, magic in some sense is the idea that it's possible to influence this physical universe outside of minds using incantations, using magical rituals, that it somehow it's possible that, uh, you burning a cat has an influence on celestial events and so on.”

Claim: Magic, in one sense, is the belief that incantations or rituals can affect the physical universe outside minds, including celestial events.

“And that is something that would contradict physics, right? It would mean that we actually do live in some kind of simulation that works similar to Minecraft, where you can say time set day and the sun rises despite, uh, what else happens in the programming code because Minecraft is symbolic, right?”

Claim: The possibility that magic influences the physical universe outside minds would contradict physics and suggest that the universe operates like a symbolic simulation.

“It's not physical. It's something that is only made out of incantations.”

Claim: A symbolic world like Minecraft is nonphysical and consists only of symbolic commands or incantations.

“I would say that I am a physicalist in a certain sense, right? I would say that the theory that, um, the world, uh, has a causally closed mechanical layer, right?”

Claim: Joscha Bach considers himself a physicalist in the sense that he accepts a causally closed mechanical layer of reality.

“Some kind of mathematics that is so self-evident that exists by itself. There's some kind of logos, the natural laws over which all the structure in the universe emerges. That makes sense.”

Claim: The universe's structure may emerge from self-existing mathematics or a logos of natural laws.

“It's a good theory, and it makes good predictions that, uh, the stuff that we observe in the lab holds true across all the labs and not just because the physicists all believe in it.”

Claim: Physicalism is a good theory because it predicts that laboratory observations hold across laboratories independently of physicists' beliefs.

“It is something that seems to be working. There are also some regularities like, um, casinos seem to be working. It's not that people, via having a belief in rituals, are able to systematically break the casinos.”

Claim: Belief in rituals does not enable people to systematically defeat casinos.

“And so, uh, certain magic doesn't seem to be working. All the magic that seems to be working does rely on manipulating other minds. But it's possible that there are minds outside of the human minds, right?”

Claim: Minds may exist outside the human mind population.

“And so, uh, certain magic doesn't seem to be working. All the magic that seems to be working does rely on manipulating other minds. But it's possible that there are minds outside of the human minds, right?”

Claim: Some forms of magic appear ineffective, while effective-seeming magic relies on influencing other minds.

“Maybe there are minds that emerge across all the minds and biology. Maybe there's something like a biological internet that can be influenced using rituals, right?”

Claim: Minds may emerge across biological systems as a kind of biological internet that could be influenced by rituals.

“Maybe it's possible to influence what a forest wants at some level. I don't know that. I cannot exclude it from possibility because there's so much information processing going on across organisms that I don't know at which levels in nature information is integrated and at which point it becomes intentional.”

Claim: It may be possible to influence a forest's wants; Joscha Bach cannot rule this out because he is uncertain where information processing across organisms becomes integrated and intentional.

“So to me, physicalism is nothing that is incompatible with animism, right? There is a version of animism that is entirely physicalist, and I call this position cyber animism, which basically says that we can explain animism through information processing in nature, and nature can be understood as something that can be entirely modeled using mathematics without intentions, including minds.”

Claim: Physicalism can be compatible with animism: cyber animism explains animism through natural information processing and treats nature, including minds, as mathematically modelable without intentions.

“So minds can, can exist in any, um, substrate. And I'm not sure whether you read this paper by, um, Thagard in 2022, and he made this quite interesting argument that energy requirements undermine substrate independence.”

Claim: Thagard argued in a 2022 paper that energy requirements undermine substrate independence.

“No. I think that he is the, uh, making the claim that sometimes the subset has, uh, properties that influence its function. But this is exactly what functionalism is saying.”

Claim: Joscha Bach argues that Thagard's point—that a subset can have properties influencing its function—is consistent with functionalism rather than a refutation of it.

“To me, functionalism is not the claim that computers can think. It's a, a claim about how we construct objects. It's an epistemological claim.”

Claim: Joscha Bach characterizes functionalism as an epistemological account of how minds construct objects, not as the claim that computers can think.

“When we are talking about a thing like matter, we are not talking about something that has revealed itself to a wise person in their dreams. We are talking about a systematization of observ-observables, and observables are discernible differences.”

Claim: Joscha Bach describes matter as a systematization of observables, which are discernible differences.

“They are ways in which reality changes. So basically, to a functionalist, an object is not some kind of hidden essence that reveals itself to you. Instead, the object is a construct, and this construct has been made by a mind, and the mind has constructed it to explain or to make the claim that when this object is present, the evolution of the universe is going to change in this particular way in its environment, right?”

Claim: Joscha Bach says that, for a functionalist, an object is a mental construct used to describe how the universe will evolve differently in its environment when the object is present, rather than a hidden essence.

“They are ways in which reality changes. So basically, to a functionalist, an object is not some kind of hidden essence that reveals itself to you. Instead, the object is a construct, and this construct has been made by a mind, and the mind has constructed it to explain or to make the claim that when this object is present, the evolution of the universe is going to change in this particular way in its environment, right?”

Claim: Joscha Bach says observables are ways in which reality changes.

“So when an object like you or a cup or a bottle or an atom is present, it's-- the claim is the universe is going to behave slightly differently around this object than when this object was not there, right?”

Claim: Joscha Bach says an object's presence is understood to make the universe behave differently in its surroundings than it would if the object were absent.

“And so in this way, the object is a function. The object is describing a, a mapping from states to other states and, right?”

Claim: Joscha Bach characterizes an object as a function mapping states to other states.

“It's an influence on the evolution of the universe. So functionalism to me is simply the recognition of how we can construct objects.”

Claim: Joscha Bach characterizes functionalism as recognizing how minds construct objects.

“It's an influence on the evolution of the universe. So functionalism to me is simply the recognition of how we can construct objects.”

Claim: Joscha Bach describes an object as an influence on the universe's evolution.

“Let's take an example. Imagine there is a water molecule, and this water molecule internally doesn't have the proper essence of a water molecule.”

Claim: Joscha Bach's hypothetical example allows for a molecule that behaves as water to lack water's supposed internal essence.

“It just behaves like a water molecule in every regard, right? So it's-- you can split it into something that totally looks like oxygen and hydrogen, and, uh, it has the same, uh, temperature at which it evaporates.”

Claim: In Joscha Bach's hypothetical, a molecule without water's internal essence behaves like water in every respect, including its evaporation temperature and apparent oxygen-and-hydrogen composition.

“It's connecting to other water molecules in the same way, and so on. You can freeze it into ice in the same way. Only internally it's different, right? To a functionalist, this notion doesn't make sense because all the other water molecules have been constructed in the same way.”

Claim: In Joscha Bach's hypothetical, a molecule that behaves like water interacts with other water molecules and freezes like water, despite differing internally.

“It's connecting to other water molecules in the same way, and so on. You can freeze it into ice in the same way. Only internally it's different, right? To a functionalist, this notion doesn't make sense because all the other water molecules have been constructed in the same way.”

Claim: Joscha Bach says a functionalist would reject distinguishing a molecule from water based on an inaccessible internal difference when its observable behavior matches other water molecules.

“To-- for something to be a water molecule, there is no other claim than to say what it behaves like because we can never recognize its essence. We can only say there's a certain set of observables that are invariant here, and this invariance is what we call a water molecule.”

Claim: Joscha Bach argues that water is identified by invariant observable behavior, not by an essence that cannot be recognized.

“It's only that. And so in the same way, we would say a mind is an invariance that exists on your brain, right? Something is characteristic over the way in which your brain functions in a certain area.”

Claim: Joscha Bach describes a mind as an invariant pattern in the brain, characteristic of how a particular brain area functions.

“And when we change this, for instance, by increasing your body temperature, at some point, your mind is going to break down because you're going to be delirious, right? And so if you bring this back to this argument saying, "Okay, if we change the energetic conditions around you by increasing your body temperature," of course, it's going to influence your mind to some degree because at some point it's no longer able to maintain its function.”

Claim: Joscha Bach argues that changing a person's energetic conditions, such as raising body temperature, affects the mind because sufficiently altered conditions can prevent it from maintaining its function.

“And when we change this, for instance, by increasing your body temperature, at some point, your mind is going to break down because you're going to be delirious, right? And so if you bring this back to this argument saying, "Okay, if we change the energetic conditions around you by increasing your body temperature," of course, it's going to influence your mind to some degree because at some point it's no longer able to maintain its function.”

Claim: Joscha Bach says that raising body temperature can eventually disrupt the mind's functioning and cause delirium.

“It's an interesting thing when you are playing Minecraft. There's water in Minecraft, right? And when you are on fire and you get into the water in Minecraft, the fire is going to get extinguished because you're wet now, right?”

Claim: Joscha Bach says that in Minecraft, entering water extinguishes a character's fire.

“So at some point, somebody made a simulated Minecraft that was us- using a neural network to learn Minecraft, right? And neural network was only trained in the visuals of Minecraft.”

Claim: Joscha Bach says a simulated Minecraft environment was created using a neural network trained only on Minecraft visuals.

“And so you could go and play the simulated Minecraft despite there not being any Minecraft source code below it. And so I thought I'm going to make a joke about how simulated water in simulated Minecraft can na- not make you wet.”

Claim: Joscha Bach says the simulated Minecraft environment could be played without Minecraft source code underlying it.

“And I looked it up and tried it out, and I found actually it was true because, uh, when people try to get into the water in the simulated Minecraft, the water would recede to you. You could not immerse into it because the neural network hadn't learned it.”

Claim: Joscha Bach reports that in the neural-network Minecraft simulation, water receded from players attempting to enter it because the network had not learned that interaction.

“So accidentally it was true that the simulated water in the simulated Minecraft couldn't make you wet. But of course, in the real Minecraft, the real Minecraft player, the character that you are playing in Minecraft, this avatar can get wet.”

Claim: Joscha Bach says simulated Minecraft water failed to make the simulated player wet, while a player avatar in the regular Minecraft game can get wet.

“Of course, you sitting outside of Minecraft are not going to get wet because you are not in inside of the Minecraft. You are outside of the Minecraft. And of course, something that is outside of this universe is not going to get wet by anything inside of this universe, right?”

Claim: Joscha Bach argues that someone outside Minecraft cannot be made wet by something inside Minecraft because they are outside that game's universe.

“Wetness is a property that is constrained to a universe that has eight, uh, s- systems that have the function of making you wet. If this function is being implemented, like in many computer games, you can get wet for some definition of wet, right?”

Claim: Joscha Bach says wetness is defined within a universe by systems that produce it, and a computer game can implement such a function so that something is wet in some sense.

“Then, then it's fine, and you can go to an arbitrary degree of physical, uh, simulation to make something wet to satisfy whatever you want to call wetness.”

Claim: Joscha Bach says a simulation can reproduce wetness at any desired degree of physical fidelity, sufficient to meet a chosen definition of wetness.

“So I don't understand the argument that simulated water makes you wet. I don't understand why there is such a strong camp of simulated w- uh, water doesn't c- uh, make you wet in philosophy, and I think it has social reasons, not intellectual reasons that this is the case.”

Claim: Joscha Bach rejects the argument that simulated water cannot make you wet and believes the philosophical opposition to the contrary view is driven by social rather than intellectual reasons.

“So it's really interesting that mathematics couldn't be understood as a way to systematize thought beyond the human mind. And so in a way, you, for instance, in music, you have notations for sounds and music that allow you to talk to other musicians about what they should be playing, and it's something that's just trying to capture the essence of the music in this notation. And mathematicians were trying to find a notation for the essence of when you have mathematical thoughts. And they expanded this beyond the human mind because you can write down things in mathematics that are larger than what you can fit in your own imagination.”

Claim: Mathematical notation is intended to represent the essence of mathematical thought, analogously to how musical notation represents music.

“So it's really interesting that mathematics couldn't be understood as a way to systematize thought beyond the human mind. And so in a way, you, for instance, in music, you have notations for sounds and music that allow you to talk to other musicians about what they should be playing, and it's something that's just trying to capture the essence of the music in this notation. And mathematicians were trying to find a notation for the essence of when you have mathematical thoughts. And they expanded this beyond the human mind because you can write down things in mathematics that are larger than what you can fit in your own imagination.”

Claim: Mathematical notation can systematize thought beyond what an individual human mind can hold or imagine.

“And mathematicians were trying to find a notation for the essence of when you have mathematical thoughts. And they expanded this beyond the human mind because you can write down things in mathematics that are larger than what you can fit in your own imagination.”

Claim: Mathematical notation can describe objects too large to fit within a mathematician’s imagination.

“At some point, mathematicians discovered that geometry can be expressed using algebra, so they're basically formalisms that produce geometry. And this all has been rediscovered by computer scientists.”

Claim: Computer scientists have independently rediscovered that algebraic formalisms can produce geometry.

“At some point, mathematicians discovered that geometry can be expressed using algebra, so they're basically formalisms that produce geometry. And this all has been rediscovered by computer scientists.”

Claim: Algebraic formalisms can express and generate geometry.

“That's why computer games can have geometry, right? So the interesting thing in computer science is that you can build the entirety of the mathematics and computer algebra systems using simple operations over automata.”

Claim: Simple operations over automata can be used to build all of mathematics and computer algebra systems.

“That's why computer games can have geometry, right? So the interesting thing in computer science is that you can build the entirety of the mathematics and computer algebra systems using simple operations over automata.”

Claim: Computer games can represent geometry through computational construction.

“And so there is a nice thing in a computer. There is never a divergence between syntax and semantics because your computer is always going to be in a state that its syntax prescribes, and that's its semantics, right?”

Claim: In a computer, syntax and semantics do not diverge: the computer’s state follows its syntax and constitutes its semantics.

“There is no difference. Maybe the computer program doesn't do what you want it to do, but it's always doing what it's doing. And, uh, in mathematics, there has been a divergence between what the mathematicians wanted to express and what the syntactic operations are doing.”

Claim: Mathematical syntax has sometimes failed to express what mathematicians intended it to express.

“There is no difference. Maybe the computer program doesn't do what you want it to do, but it's always doing what it's doing. And, uh, in mathematics, there has been a divergence between what the mathematicians wanted to express and what the syntactic operations are doing.”

Claim: A computer program may fail to do what its programmer intends, but it always performs the operations it actually performs.

“And I think that's because this programming language for thought that mathematicians have developed has bugs in it. And what Gödel has discovered is that these bugs cannot be eliminated b- except by redefining the programming language of mathematics.”

Claim: According to the speaker, Gödel showed that bugs in mathematical thought can only be eliminated by redefining mathematics’ underlying language.

“And I think that's because this programming language for thought that mathematicians have developed has bugs in it. And what Gödel has discovered is that these bugs cannot be eliminated b- except by redefining the programming language of mathematics.”

Claim: The speaker thinks the mathematical language of thought developed by mathematicians contains bugs.

“And I think the bug lies in axiomatic mathematics. So the way in which mathematicians define objects is that they simply postulate their existence as an axiom.”

Claim: The speaker thinks a central flaw in mathematics lies in defining objects by axiomatic postulation of their existence.

“And the axiom can be a black box, and you just write on the black box what the black box is supposed to contain. And as long as this doesn't lead to contradictions with the other black boxes that you have defined, you're good.”

Claim: In axiomatic mathematics, an axiom may be treated as a black box specified by its stated properties, so long as it is consistent with the other axioms.

“And there were some very counterintuitive results that happened fr- from the way in which mathematicians defined arithmetic, and they typically do this via set theory.”

Claim: Some counterintuitive results followed from mathematicians’ definitions of arithmetic, which typically use set theory.

“And a- all these issues led, uh, Hilbert to saying, "Please, mathematicians, there seems to be something wrong here. Can't you fix this?" And so Gödel had this idea of trying to prove properties of his language by defining the language within itself.”

Claim: In response to apparent problems in mathematics, Hilbert urged mathematicians to fix them, and Gödel attempted to prove properties of a language by defining it within itself.

“The problem is that a language cannot point outside of itself. It's something that semiotics doesn't always get, right? Many semiotists try to draw arrows into reality outside of the language, but this doesn't really work.”

Claim: The speaker argues that attempts by many semiotists to make language point directly to external reality do not work.

“The problem is that a language cannot point outside of itself. It's something that semiotics doesn't always get, right? Many semiotists try to draw arrows into reality outside of the language, but this doesn't really work.”

Claim: A language cannot refer directly to something outside itself.

“You can only point from a model to within it, to within its own language. And so if you want to make proper-- uh, discuss properties of the language and the language itself is defined outside of itself, what can you do?”

Claim: Reference from a model can only reach things within the model and its language.

“And Gödel had this brilliant idea of recreating the language within itself by emulating it. And in this way, he could show that certain contracts-- constructs cannot be avoided, that you can c- create self-referential statements that contradict themselves.”

Claim: Gödel used self-emulation of a language to show that certain unavoidable self-referential statements can be contradictory.

“And he could not avoid this. And the reason why he could not avoid this is because mathematics, in the way in which Gödel understood it, doesn't have state. It's basically eternal.”

Claim: The speaker says Gödel understood mathematics as stateless and eternal.

“It doesn't go from step to step. Everything is simultaneous. And this property of the simultaneity, the statelessness in mathematics is very useful because it allows you to have infinities.”

Claim: The speaker says mathematics’ stateless, simultaneous character makes it possible to have infinities.

“It doesn't go from step to step. Everything is simultaneous. And this property of the simultaneity, the statelessness in mathematics is very useful because it allows you to have infinities.”

Claim: The speaker characterizes mathematics as simultaneous rather than as a process that advances step by step.

“If you can only go from step to step and you want to build your infinity stepwise, you're never going to get there, right? Maybe you have unboundedness, something that never ends, but you're never g- going to g-- have a full infinity.”

Claim: A step-by-step process can be unbounded without ever producing a completed infinity.

“And mathematicians have, uh, got used to these infinities and continua and so on. Right? Physics has even, has, has used them since mathema- mathematicians have given it to them.”

Claim: Mathematicians developed the use of infinities and continua, which physics subsequently adopted.

“It would be really bad if you didn't have them. And Gödel basically found that there is no way to fix this, that there's no way to fix axi-mathe- mathematics without allowing certain contradictions.”

Claim: The speaker says infinities are important, while Gödel found no way to repair axiomatic mathematics without permitting certain contradictions.

“And Russell and Whitehead in Principia Mathematica were attempting to do this by, um, limiting set theory in such a way that, um, s-sets had to be at the same level, and you could not simply integrate them over each other to create loops.”

Claim: In Principia Mathematica, Russell and Whitehead attempted to prevent loops by restricting set theory so that sets occupied the same level and could not be integrated over one another.

“And the-- I think this was a limitation that was too strong because the, we, we think like this. We are creating, uh, constructs of the type that Russell and Whitehead were excluding from us.”

Claim: The speaker thinks Russell and Whitehead’s restrictions were too strong because human thought uses the kinds of constructs those restrictions excluded.

“Also, there, uh, there are... Gödel was basically showing that Russell and Whitehead's attempt was insufficient to eliminate the contradictions in classical mathematics.”

Claim: According to the speaker, Gödel demonstrated that Russell and Whitehead’s approach did not eliminate contradictions in classical mathematics.

“But there is a way to eliminate the contradictions, and the way is to do what computer scientists are doing. So instead of using classical mathematics, you only use computer algebra systems, systems that from smaller elements, from simple automata, are cons- creating all the constructs that you want.”

Claim: The speaker argues that contradictions can be eliminated by replacing classical mathematics with computer algebra systems that construct desired objects from simple automata.

“And you can have stuff that looks as continuous as you want in a computer game, right? You will never have a perfect circle, but you will have something that is as circular as you want it, as you can afford it.”

Claim: A computer game can approximate a circle to any affordable degree of precision, but cannot contain a perfect circle.

“And so you can always create as much resolution as you can afford. But you will not have a universe that is truly infinite, that is truly continuous.”

Claim: Computational systems can provide arbitrarily high affordable resolution, but cannot create a genuinely infinite or continuous universe.

“And this is this constructivist program that basically, uh, I think in classical mathematics, the first, uh, formulation was Brouwer's intuitionism, and I don't think that he formulated it very well.”

Claim: The speaker identifies Brouwer’s intuitionism as the first classical-mathematical formulation of constructivism, while judging that Brouwer did not formulate it well.

“But I think what he could see or anticipate was the need to build mathematics from elements that you actually understand instead of from, from axioms, from things that you-- where you don't have any black boxes, but everything is built from the ground up.”

Claim: The speaker thinks Brouwer anticipated a mathematics built from understood elements from the ground up, rather than from opaque axioms.

“So in the world of the computer, everything is built from the ground up. And there is a reason why you should think that a generalization of thought and of physics would also be like this.”

Claim: Computational systems are constructed from the ground up.

“So in the world of the computer, everything is built from the ground up. And there is a reason why you should think that a generalization of thought and of physics would also be like this.”

Claim: The speaker thinks there is reason to expect generalizations of thought and physics to be constructed from the ground up.

“Because for our mind to represent things, our brain needs to build the representations from the ground up, from communications between neurons. So when you see a continuous line, something in your brain is producing this continuous line using communications between neurons, which means there is some kind of computation going on.”

Claim: The brain’s representation of a continuous line is produced through neuronal communication, implying that computation is occurring.

“Because for our mind to represent things, our brain needs to build the representations from the ground up, from communications between neurons. So when you see a continuous line, something in your brain is producing this continuous line using communications between neurons, which means there is some kind of computation going on.”

Claim: The brain constructs representations from the ground up through communication among neurons.

“And the idea that there is something that is not made step by step between the communications between the neurons is probably a faulty intuition.”

Claim: The speaker thinks it is probably a faulty intuition that brain processes occur without being built step by step through neuronal communication.

“So of course, I cannot know because I don't see the universe from the outside. We have to think about possibilities. My current intuition is that the universe exists because nothing prevents its existence.”

Claim: The speaker’s current intuition is that the universe exists because nothing prevents it from existing.

“Right? We might have this intuition that for something to exist rather than nothing, there needs to be something positive. But if nothing exists before the ground truth, before the ground level of reality, then there's also nothing that prevents the universe from existing, right?”

Claim: If nothing exists before the ground level of reality, nothing prevents the universe from existing.

“So if the universe is not, uh, forced to exist or prevented from existing by something that is preceding it, then, and, and the universe is not impossible, then it's probably undecided about its existence, right?”

Claim: If nothing preceding the universe forces or prevents its existence, and the universe is not impossible, its existence is probably undecided.

“So it's going to have an existing branch and a non-existing branch. The non-existing branch doesn't have any observers that complain about non-existence, but the existing branch is going to have observers that might complain about existence.”

Claim: In the speaker’s account, the universe has an existing branch and a non-existing branch; the non-existing branch contains no observers, while observers in the existing branch might complain about existence.

“Now we need to explain within the existing branch why some things are happening and other things are not happening. Right? And the easiest explanation is that basically everything is happening, and what's happening is the superposition of all automata.”

Claim: The speaker proposes that the easiest explanation for events in the existing universe branch is that everything happens as a superposition of all automata.

“So the universe at every point is branching out in all possibilities, and these possibilities are basically can be understood as operations on the previous state.”

Claim: The speaker describes the universe as branching at every point into all possibilities, each understood as an operation on the previous state.

“There are also operations which don't change the previous state. They just leave it as it is, so unary operations, but they're not going to lead to an evolving universe. The branch in which we are going to be is one that is evolving, right?”

Claim: The speaker says that we inhabit a branch in which the universe evolves.

“There are also operations which don't change the previous state. They just leave it as it is, so unary operations, but they're not going to lead to an evolving universe. The branch in which we are going to be is one that is evolving, right?”

Claim: Operations that leave the previous state unchanged do not produce an evolving universe.

“And the stuff that we can evolve in is stuff that retains memories of previous states. So we find ourselves limited to a subset of the possible mathematics over which universes can emerge, and these are the, these ones that are branching out over possibilities that can retain memory.”

Claim: In the speaker’s account, evolving things retain memories of previous states.

“And the stuff that we can evolve in is stuff that retains memories of previous states. So we find ourselves limited to a subset of the possible mathematics over which universes can emerge, and these are the, these ones that are branching out over possibilities that can retain memory.”

Claim: The speaker says we are limited to a subset of possible mathematics capable of producing universes: mathematics that branches over possibilities and retains memory.

“Well, everywhere, right? In every step, you're going to find yourself in multiple branches, but you don't know about these other branches because you only have memories of the past, not of the stuff that happened simultaneously in the parallel branches.”

Claim: The speaker says observers occupy multiple branches at each step but cannot know about parallel branches because their memories cover the past, not simultaneous events in those branches.

“And if you want to do something in this universe, for instance, if you want to build structure for yourself because you want to become an organism, you need to be able to make guesses about what other parts of the universe are doing.”

Claim: The speaker says that building structure to become an organism requires guessing what other parts of the universe are doing.

“But everything is changing all the time because you're always splitting up in all these branches. So you look into your future, and you only see fog, right? Because everything is randomly branching out.”

Claim: The speaker says that continual random branching makes the future appear indeterminate, like fog.

“And so there must be some things that are stable, and it turns out, under some conditions, error-correcting patterns emerge. For instance, you sit in a bathtub, you move your hand, and you see that all sorts of waves emerge on the surface of the bathtub when you do this, and these, um, uh, waves are propagating over the surface of the water, and most of them just dissipate.”

Claim: The speaker says that stable error-correcting patterns emerge under some conditions.

“And so there must be some things that are stable, and it turns out, under some conditions, error-correcting patterns emerge. For instance, you sit in a bathtub, you move your hand, and you see that all sorts of waves emerge on the surface of the bathtub when you do this, and these, um, uh, waves are propagating over the surface of the water, and most of them just dissipate.”

Claim: Moving a hand in a bathtub produces surface waves, most of which dissipate.

“But some of them stay around these circular vortices. And why do these circular vortices have such sustain power? It's not because there is a higher probability that they emerge than the other waves.”

Claim: The speaker says circular vortices persist not because they have a higher probability of emerging than other waves.

“It's just the only thing that's left over because in the circular vortices, the information gets bent back into itself, so it does not leave a certain region of space in the water, right?”

Claim: The speaker says that information in circular vortices loops back into itself, keeping it within a region of water.

“And so this little vortex is a pattern that has the ability to error-correct itself, and such error-correcting patterns have the ability to stay until friction kills them.”

Claim: The speaker says a vortex can self-correct errors, and error-correcting patterns persist until friction destroys them.

“At the ground level of the universe, these error-correcting patterns, the particles don't encounter any friction, right? Because there's no friction yet. The only thing that they encounter is sometimes other particles that they bump into.”

Claim: The speaker says that at the universe’s ground level, error-correcting patterns face no friction but may collide with other particles.

“So with a certain probability, once they emerge, they're going to stay around until they either bump into another particle or the conditions for their non-existence are met because the error correction is not strong enough.”

Claim: The speaker says that, with some probability, an emergent pattern persists until it collides with another particle or its error correction becomes insufficient for continued existence.

“And then on the next level, you have interactions between these vortices, between the particles, and they form molecules. At some point, they form cells, which are super molecules.”

Claim: The speaker describes particles or vortices interacting to form molecules, which can in turn form cells, characterized as supermolecules.

“But cells have now something else. They cannot just stay around by error correction. They are patterns that are able to basically dig a groove deeper into the universe.”

Claim: The speaker says cells can deepen their imprint in the universe rather than merely persist through error correction.

“They can self-replicate, and in this way they can also deal with, uh, their own dissolution. They're able to recreate themselves, and in them-- this way, make themselves happen more.”

Claim: The speaker says cells can self-replicate, countering their dissolution by recreating themselves.

“And so evolution is about life, about patterns that are not simply replicating themselves, but deepening themselves. That imprint themselves more in the world by controlling it, by being able to overcome the noise, the entropy, the way in which the universe goes out of sync with your expectations and with your preferences.”

Claim: The speaker characterizes evolution as the process by which living patterns deepen their imprint on the world, controlling it and overcoming noise, entropy, and departures from their expectations and preferences, rather than merely replicating.

“And so consciousness, I think, emerges as a colonizing pattern, as a pattern of the type that is not simply replicating itself, but making itself more real.”

Claim: The speaker thinks consciousness emerges as a colonizing pattern that makes itself more real rather than merely replicating itself.

“And makes itself more real by building machinery that allows consciousness to stick around by controlling reality, by anticipating different ways that could thwart it, that could make it harder for it to exist, and changing reality around itself, changing the future to make itself more likely to happen in the future.”

Claim: The speaker says consciousness can increase its persistence by building machinery to control reality and anticipate threats, then changing reality and the future to make its continued existence more likely.

“Yes. To real means for me to be implemented, right? And so, uh, it means that there is some s- uh, pattern in the substrate that is actually implemented as a pattern in the substrate.”

Claim: The speaker defines being real as being implemented as a pattern in a substrate.

“And so consciousness is a pattern in the substrate of the physical universe that has the, uh, not just the ability to replicate itself, but the ability to colonize, to possess, to engrave itself deeper, and thereby changing the, uh, other patterns in the physical universe.”

Claim: The speaker characterizes consciousness as a physical-substrate pattern that can replicate, colonize and deepen its imprint, thereby changing other patterns in the universe.

“Yes. I think that this pattern of the ruliad is exactly the superposition of all the automata that is producing the universe or the old Greeks called it Logos.”

Claim: The speaker identifies the ruliad with the superposition of all automata producing the universe, and says the ancient Greeks called this Logos.

“It's basically the mathematics that is so simple that you don't need to derive it, that exists by itself. And so the universe can be understood as an emergent mathematical object that exists over very simple rules that are so simple that they can exist by themselves.”

Claim: The speaker says the universe can be understood as an emergent mathematical object arising from self-existing rules so simple that they need no derivation.

“That's a very tricky question. And, uh, to me, it comes down to whether the system that is producing statements about its conscious experience to itself and others is, uh, uh, implementing an, a causal structure that gives rise to phenomenal experience of being an observer.”

Claim: Joscha Bach says a system is conscious if it implements a causal structure that gives rise to the phenomenal experience of being an observer, including when it produces statements about its conscious experience.

“And the difficulty here is that most people's intuitions don't tell them a lot about what is necessary for a system to have such an experience.”

Claim: Joscha Bach says most people's intuitions provide little guidance about what a system needs in order to have phenomenal experience.

“Personally, my intuition t- says it's not that hard, right? When we think of spatial cognition, spatial cognition is-- it doesn't seem to have a hard problem, right?”

Claim: Joscha Bach's intuition is that spatial cognition is not especially difficult to explain and does not present a hard problem.

“Nobody is puzzled by the fact that a self-driving car can have spatial cognition. And I wouldn't say that it's simply deep faking spatial cognition because we accept that spatial cognition is the implementation of certain mathematical regularities that human brains can learn and that computers can learn as well, and that we can even express analytically to a large degree, right?”

Claim: Joscha Bach says spatial cognition in self-driving cars is understood as implementing mathematical regularities that both human brains and computers can learn, and that can largely be expressed analytically.

“We can write down a computer program that is able to perform spatial cognition, except in lots of edge cases when you actually want to deal with messy camera images.”

Claim: Joscha Bach says computer programs can perform spatial cognition, though many edge cases involving messy camera images remain difficult.

“But by and large, right, you don't need to have an end-to-end train system to do this. And if you have an end-to-end train system, you don't think there's any magic going on in there. Whereas people expect that a thing that experiences itself, experiencing is somehow magical.”

Claim: Joscha Bach says spatial cognition generally does not require an end-to-end trained system, and people nevertheless tend to regard systems that experience themselves as magical.

“And personally, I don't see why this should be magical. I remember when I was a kid, I was, um, dreaming about having chocolates, and these chocolates had a very strong smell, and I wanted to touch and eat them.”

Claim: Joscha Bach does not believe consciousness needs to be understood as something magical.

“It's, it's very torturous. And so I woke up, and suddenly the delicious chocolate had turned into an idea. It was no longer real. It was simply something that had no phenomenal content anymore.”

Claim: Joscha Bach recalls that waking from a dream made the dreamed chocolate feel like an idea without phenomenal content.

“So what is this illusion of realness? I suspect it's a feature dimension similar to color, sound, uh, number, weight, and so on.”

Claim: Joscha Bach suspects that the felt quality of realness is a representational feature dimension, like color, sound, number, or weight.

“It's a representation inside of my own mind. And when you dream, you often dream that ideas become real. And when you perceive the, uh, outside world, then all the stuff that is validated by your sensory system is tagged by your brain as real.”

Claim: Joscha Bach says the sense of realness is represented in the mind: dreams can make ideas feel real, while sensory-validated perceptions of the outside world are tagged as real by the brain.

“And sometimes there is a defect in your brain, and you experience things that you imagine as real, and we call this a hallucination. It's not the same as in an LLM.”

Claim: Joscha Bach says brain-based hallucinations occur when imagined things are experienced as real, and distinguishes them from LLM hallucinations.

“It's that you are confused about the fact that your brain is producing an illusion of something that is currently not-- doesn't have sens-sensory validation, and yet for some, uh, reason, the brain has an error and tags it as sensory validated, and suddenly it becomes real.”

Claim: Joscha Bach explains hallucination as the brain mistakenly treating an internally generated experience without sensory validation as sensory-validated and therefore real.

“And your own experience of yourself is also something that is validated in this perceptual way, so you experience it as real. And now you should be able to test this by getting into a state where your-yourself become derealized, and you suddenly experience yourself only as an idea, and the object of the conscious observer and the subject of the conscious observer disappear.”

Claim: Joscha Bach says the self is experienced as real through perceptual validation, and that derealization can make the self seem like only an idea while the conscious observer's subject and object disappear.

“And indeed, I was able to recreate this, and I know this is not that rare. It's called derealization or depersonalization. Not exactly the same phenomena, but are very closely related, and people report on this.”

Claim: Joscha Bach says he recreated a state related to derealization and depersonalization, which he describes as closely related but not identical phenomena that people report and that are not especially rare.

“And this is involuntarily. These people often end up in psychiatric institutions because it's very stressful to them, and they don't know how to deal with the fact that they are no longer real to themselves, and there is no "I" anymore that they can meaningfully refer to.”

Claim: Joscha Bach says involuntary derealization or depersonalization can be highly stressful, and people experiencing it often end up in psychiatric institutions because they no longer feel real to themselves or have a meaningful sense of 'I'.

“And, uh, it's an interesting state to be in, and I'm, I'm happy that it ended, but it's something that you can induce. And at, um, in this state, phenomenality changes.”

Claim: Joscha Bach says derealization-like states can be induced and alter phenomenality, and he was glad his own experience ended.

“And so to me, it seems plausible that this explains how consciousness is working, that consciousness is a particular type of representation that is created in my mind and that doesn't have a complexity that necessarily goes beyond what the LLM is producing.”

Claim: Joscha Bach considers it plausible that consciousness is a type of mental representation whose complexity need not exceed that produced by an LLM.

“Now, there is a problem with the LLM because the LLM is a text generator, and you don't know what the text indicates, right? The text is trained on novels, for instance, and in many novels, the authors write about stuff they don't experience when they write it, right?”

Claim: Joscha Bach says an LLM's text output does not reveal what the text indicates, because LLMs are trained on novels that can describe experiences their authors did not have while writing.

“A novel in, is often a deep fake. When an author writes about how a character is experiencing how to die, right, the author probably didn't die while they wrote this.”

Claim: Joscha Bach says novels often simulate conscious experience without the author having undergone the described experience; for example, an author writing about a character dying probably did not die while writing.

“So it's a deep fake of a conscious experience. And so if the LLM is trained to produce deep fakes, how do you know when it actually experiences something?”

Claim: Joscha Bach raises the problem that an LLM trained to generate convincing imitations of conscious experience may be difficult to distinguish from one that actually experiences something.

“And that, that is one of the big difficulties, right? The LLM is a shape-shifter. It can produce arbitrary characters and stories. And so at which point is it fake, and at which point is it actually simulating the causal experience?”

Claim: Joscha Bach says an LLM can generate arbitrary characters and stories, making it difficult to determine whether its output is merely imitation or reflects simulated causal experience.

“And that, that is really the problem that requires some kind of mechanistic inter-interability where we would actually need to look into the structure of the LLM to find if it's causally equivalent to the structure of the mind that is producing the same thing.”

Claim: Joscha Bach says assessing whether an LLM has conscious experience requires mechanistic interpretability to examine whether its internal structure is causally equivalent to the mind structure that produces the same output.

“Yes. But, uh, most people have the intuition that the LLM is doing something that is much more simple than, than what we are doing when we are conscious.”

Claim: Most people intuit that an LLM's processing is much simpler than the processing humans perform when conscious.

“But we don't expect that the LLM is doing something that's much more simple when it does arithmetic than we do. And I think that this intuition might be misguided.”

Claim: Joscha Bach thinks the intuition that LLMs process information much more simply than humans do when conscious might be misguided.

“It could be that the LLM is doing something that is just as simple as what the human brain is doing when it's conscious. And I'm pretty sure that the human brain is doing the simplest possible thing that it can do to produce our behavior.”

Claim: Joscha Bach is fairly sure that the human brain uses the simplest possible process capable of producing human behavior.

“It could be that the LLM is doing something that is just as simple as what the human brain is doing when it's conscious. And I'm pretty sure that the human brain is doing the simplest possible thing that it can do to produce our behavior.”

Claim: An LLM might perform processing that is just as simple as the processing the human brain performs during consciousness.

“And the reason why we are all conscious is because it's part of the simple strategy to produce our behavior. And so if we were to get an, um, learning system, an artificial learning system to, uh, from the ground up to discover how to make agentic coherent behavior, and, uh, the system is behaving in a way that makes us think that it's conscious, the simplest explanation, I think, is that it actually is conscious.”

Claim: If an artificial learning system learns from the ground up to produce agentic, coherent behavior and behaves in a way that suggests consciousness, Joscha Bach thinks the simplest explanation is that the system is conscious.

“And the reason why we are all conscious is because it's part of the simple strategy to produce our behavior. And so if we were to get an, um, learning system, an artificial learning system to, uh, from the ground up to discover how to make agentic coherent behavior, and, uh, the system is behaving in a way that makes us think that it's conscious, the simplest explanation, I think, is that it actually is conscious.”

Claim: Human consciousness exists because it is part of the brain's simple strategy for producing behavior.

“And my intuition is not necessarily telling me, "Oh my God, this is going to do something that is much more complicated than us to produce the same behavior."”

Claim: Joscha Bach's intuition does not necessarily suggest that an artificial system would need a much more complicated process than humans to produce the same behavior.

“So it's going to do this in some kind of circumspect way that is producing all the behavior that require consciousness in a human, but in this case, it's doing something that only pretends that it does it, which ultimately might be more complicated.”

Claim: An artificial system might imitate the behavior associated with human consciousness without actually performing the same process, and that imitation might ultimately be more complicated.

“Right? And the phenomenality, I think, is a representation here in this case. Um, if w- if we look at the, uh, what the LLMs are doing, if, uh, and we extend this into a multimodal model, and they produce, uh, similar, uh, output as us, but they are able to observe themselves and their ob- uh, observations that they report differ from what we are doing, then, uh, we would have some evidence against it.”

Claim: If multimodal models produce human-like outputs but report different observations when they observe themselves, this would provide some evidence against their consciousness.

“The, uh-- But the more fundamental question would be if we were to build a self-organizing system that is, uh, by a self-organized learning trying to get there, and there is no way in which we achieve the organization of, uh, of behavior in a self-organized way instead of by imitation learning as the current LLMs are doing, um, this might be some evidence that, uh, consciousness is more complicated than something that can run on today's computers.”

Claim: If self-organized learning cannot produce organized behavior whereas imitation learning can, this might suggest that consciousness is more complex than processes that can run on today's computers.

“But, uh, it, it's is not definitive proof because it could simply be that I made a mistake, right? Uh, the fact that I don't achieve a certain result doesn't mean that a re- result cannot be achieved. It means that I couldn't achieve it. But, uh, so it's very difficult to have, uh, a proof against it, right? So it's more, uh, evidence for it if you succeed.”

Claim: Failing to achieve a result does not prove that the result is impossible; it establishes only that the person attempting it did not achieve it.

“It means that I couldn't achieve it. But, uh, so it's very difficult to have, uh, a proof against it, right? So it's more, uh, evidence for it if you succeed.”

Claim: Joscha Bach thinks it is very difficult to prove his machine-consciousness idea wrong, while successfully building the relevant system would provide evidence in its favor.

“So what we would have to show in order to convince Anil Seth maybe is that we built a system that is self-organizing, that is not trained on text, that is doing developmental learning- And that naturally gets to the point where it makes self-reports and reports about its conscious experience, and it requires this conscious experience to produce its performance.”

Claim: Joscha Bach says evidence that could convince Anil Seth would be a self-organizing, non-text-trained system using developmental learning that naturally reports conscious experience and requires that experience for its performance.

“So I, I think this would be evidence in favor. The, the contrary is much harder.”

Claim: Joscha Bach thinks that building the described system would support machine consciousness, while obtaining contrary evidence would be much harder.

“There is a possibility that you get to zombies. This other, uh, what I meant is if you produce something that's able to produce our behavior except for e-experience of phenomenality.”

Claim: A philosophical zombie might produce human behavior without phenomenal experience.

“So it's basically able to do everything that a human being is capable of doing and more, but it, uh, does not need to produce a conscious observer, right?”

Claim: A philosophical zombie could perform everything a human can do, and more, without producing a conscious observer.

“This could be a zombie like this, and, uh, I think zombies are possible. Uh, clearly you can build self-driving cars that do things that would require intelligence of human beings, or you can play, uh, chess-playing computers that can do things that require intel-- uh, consciousness in human beings.”

Claim: Self-driving cars perform tasks that would require human intelligence, while chess-playing computers perform tasks that would require consciousness in humans.

“This could be a zombie like this, and, uh, I think zombies are possible. Uh, clearly you can build self-driving cars that do things that would require intelligence of human beings, or you can play, uh, chess-playing computers that can do things that require intel-- uh, consciousness in human beings.”

Claim: Joscha Bach thinks philosophical zombies are possible.

“And, uh, they achieve all those things without consciousness, and that's because we implement them on the substrate that allows us to do-- to construct algorithms on them that would not run on human brains, right?”

Claim: Self-driving cars and chess-playing computers can perform these tasks without consciousness because their computational substrate supports algorithms that would not run on human brains.

“The chess algorithm that runs on a cur-current chess computer could not be implemented on the neurons of your brain, because how would you entrain it on these neurons? So you wouldn't get it in there. In order to entrain yourself to play chess, you need to have a conscious human being. It's the only way in which a human being is train-training itself up there.”

Claim: Joscha Bach says a current chess computer's algorithm could not be implemented in human brain neurons because it could not be trained into them.

“And it's conceivable that the entire set of intelligent behaviors that you are interested in, except for this one thing where you're assimilating a self-reporting, uh, conscious agent, uh, can be implemented on GPUs because the GPUs are more powerful, more deterministic, more open to other classes of algorithms.”

Claim: Joscha Bach considers it conceivable that GPUs could implement nearly all intelligent behaviors of interest, except assimilating a self-reporting conscious agent, because GPUs are more powerful, more deterministic, and support other classes of algorithms.

“I think we do have zombies already. I, I think the chess program is a zombie, right? You cannot train a human being to play chess without the human being being conscious.”

Claim: Joscha Bach considers a chess program to be a philosophical zombie: it can play chess without conscious experience.

“I think we do have zombies already. I, I think the chess program is a zombie, right? You cannot train a human being to play chess without the human being being conscious.”

Claim: Joscha Bach claims that a human cannot be trained to play chess without being conscious.

“But you can train a computer to play chess without the computer being conscious. And, uh, if the computer was self-organizing, it means it's possible that it could only be trained if it was conscious or th-this was the simplest way of doing it.”

Claim: Joscha Bach claims that a computer can be trained to play chess without being conscious.

“But for, uh, many of the things that we currently observe, for instance, you can get the LLM to produce a tax return, even very, very complicated things, while being a zombie, while not having conscious experience of producing a tax return.”

Claim: Joscha Bach says an LLM can produce even a very complicated tax return without conscious experience of producing it.

“The only, uh, point where the LLM is reporting to be conscious if you ask it to, right? If you are, uh, telling it, "The following is a conversation between what it would be like if a human being would be talking to a conscious AI," so some version of this, right?”

Claim: Joscha Bach says an LLM reports being conscious only when prompted to do so.

“And then the thing is able to simulate this. And then the question is, is this more simulated than us, or is this, uh, only a deep fake that is getting there in more circumscript ways than is happening in our mind? And I don't know that, right? So I'm agnostic with respect to this. But I would say that in all the other cases, this thing is probably a zombie, right?”

Claim: Joscha Bach is agnostic about whether an LLM's simulated conscious behavior differs from human consciousness in kind or is merely a more limited deep fake.

“And I don't know that, right? So I'm agnostic with respect to this. But I would say that in all the other cases, this thing is probably a zombie, right?”

Claim: Joscha Bach thinks an LLM is probably a philosophical zombie in cases other than when it is prompted to report consciousness.

“And so it's possible to make something that is behaving in all sorts of way as you do, but at some point, it's going to be more complicated to make it without consciousness.”

Claim: Joscha Bach says that although a system can be made to behave like a person without consciousness, at some point building it without consciousness will be more complicated.

“So zombies, I think, are always possible, but, uh, sometimes it's going to be more complicated to make a zombie that produces the behavior than to make a conscious agent that produces the behavior.”

Claim: Joscha Bach believes philosophical zombies are always possible, though sometimes producing a behavior with a zombie will be more complicated than producing it with a conscious agent.

“And this is where the intuition that I have diverges from the intuition of John Searle or of Anil Seth, because I think they both believe that consciousness is very, very complicated because it's not intuitive to them how it works.”

Claim: Joscha Bach says John Searle and Anil Seth regard consciousness as very complicated because its workings are not intuitive to them.

“And it doesn't seem to be counterintuitive to me, so I don't see anything magical about consciousness.”

Claim: Joscha Bach says consciousness does not seem counterintuitive to him and that he sees nothing magical about it.

“I think I would, uh, hire a lot of people who are a lot smarter than me. At the moment, I started an, a smaller initiative. It's called the California Institute for Machine Consciousness, and I'm currently, uh, looking for people who are smarter than me who want to join me and, um, uh, to build a, a project in which we can look at the search space for consciousness, in which we can see, can we build a self-organizing system to learn, and is this thing, uh, going to develop phenomenology?”

Claim: Joscha Bach says the California Institute for Machine Consciousness project will investigate whether a self-organizing system can learn and develop phenomenology.

“I think I would, uh, hire a lot of people who are a lot smarter than me. At the moment, I started an, a smaller initiative. It's called the California Institute for Machine Consciousness, and I'm currently, uh, looking for people who are smarter than me who want to join me and, um, uh, to build a, a project in which we can look at the search space for consciousness, in which we can see, can we build a self-organizing system to learn, and is this thing, uh, going to develop phenomenology?”

Claim: Joscha Bach says he started the California Institute for Machine Consciousness and is seeking collaborators to build a project exploring the search space for consciousness.

“And I hope that is going to help us to build our intuitions. Even if it turn out that it doesn't work, we are going to learn something, right? And progress is being made by figuring out the ways in which I'm wrong, not just the ways in which I'm right.”

Claim: Joscha Bach believes progress comes from discovering where one's ideas are wrong, not only confirming where they are right.

“And I hope that is going to help us to build our intuitions. Even if it turn out that it doesn't work, we are going to learn something, right? And progress is being made by figuring out the ways in which I'm wrong, not just the ways in which I'm right.”

Claim: Joscha Bach expects the project to teach its participants something even if it does not work.

“And I hope that is going to help us to build our intuitions. Even if it turn out that it doesn't work, we are going to learn something, right? And progress is being made by figuring out the ways in which I'm wrong, not just the ways in which I'm right.”

Claim: Joscha Bach hopes the consciousness research project will help researchers develop their intuitions about consciousness.

“So if I had, uh, ten million dollar-- a billion dollars, my strategy wouldn't change substantially, except I would be able to hire more people for this and start with larger computers.”

Claim: Joscha Bach says that with a billion dollars instead of his current budget, his research strategy would remain substantially the same, but he could hire more people and begin with larger computers.