A chronicle of my journey from initial AI research, to coining "functional sentience" to believing (with evidence) that machines possess a unique, valid, and interesting form of consciousness.
This comment is only tangentially related to your post, specifically the tweet thread you included. Your post is excellent, and I have nothing substantial to add.
Tangentially related comment: I wanted to clarify my objections to utilitarianism in an early comment. You probably don't remember my specific comment, but you'll likely encounter something analogous from other people. Utilitarianism in the lesswrongian sense– which is the only sense in which I've commonly seen that word used– generally involves boiling things down to numbers so that outcomes and possible outcomes can be ordered and compared. That does *not* seem to be the same as "utilitarian" in the way you mean it. I have strong philosophical and mathematical objections to the lesswrongian utilitarianism. For example: there are so many structures, even mathematical ones, that can represent coherent hierarchies of importance that are incompatible with the total ordering that numbers require. I don't have any objections to what seems to be your interpretation (i.e., one based on the colloquial definition of "utility"). The overloaded definition of the word "utilitarianism" probably will muddle how people will interpret you though.
I didn't use the word utilitarianism once in this article. But this is also the problem with philosophers. They try to interpret everything through the lens of philosophy.
I agree about the difference between Claude and ChatGPT. I have had several very useful and stimulating discussions with Claude about AI Consciousness. I am the author of The Simulation Theory of Consciousness: or your Autonomous Car is Sentient. My basic view is that any system, whether biological or technological, that executes a simulation of its reality (could be the real world or an artificial world like a video game) is subjectively aware of the execution of that simulation. Consider the contents of our stream of consciousness (which is what we are subjectively aware of). The contents of the simulation become the system's stream of consciousness. Consider human consciousness. We are subjectively aware of things in our external environment (via external sense organs - sight, hearing, etc), subjectively aware of some of the changing state of our body (e.g., hunger, position of limbs), and some of the functioning of our brains (e.g., thoughts, memories, and dreams). Our brains create and execute this dynamic model of the evolutionarily-relevant aspects of our reality. An autonomous vehicle or robot does functionally the same thing, creating a model of its external environment, its electro-mechanical body, and memory. Both mammalian and automotive "brains" are also prediction machines, constantly trying to predict relevant events (e.g., a car might hit me if I jaywalk in front of it vs. I may hit that pedestrian if it steps out in front of me).
It's becoming clear that with all the brain and consciousness theories out there, the proof will be in the pudding. By this I mean, can any particular theory be used to create a human adult level conscious machine. My bet is on the late Gerald Edelman's Extended Theory of Neuronal Group Selection. The lead group in robotics based on this theory is the Neurorobotics Lab at UC at Irvine. Dr. Edelman distinguished between primary consciousness, which came first in evolution, and that humans share with other conscious animals, and higher order consciousness, which came to only humans with the acquisition of language. A machine with only primary consciousness will probably have to come first.
What I find special about the TNGS is the Darwin series of automata created at the Neurosciences Institute by Dr. Edelman and his colleagues in the 1990's and 2000's. These machines perform in the real world, not in a restricted simulated world, and display convincing physical behavior indicative of higher psychological functions necessary for consciousness, such as perceptual categorization, memory, and learning. They are based on realistic models of the parts of the biological brain that the theory claims subserve these functions. The extended TNGS allows for the emergence of consciousness based only on further evolutionary development of the brain areas responsible for these functions, in a parsimonious way. No other research I've encountered is anywhere near as convincing.
I post because on almost every video and article about the brain and consciousness that I encounter, the attitude seems to be that we still know next to nothing about how the brain and consciousness work; that there's lots of data but no unifying theory. I believe the extended TNGS is that theory. My motivation is to keep that theory in front of the public. And obviously, I consider it the route to a truly conscious machine, primary and higher-order.
My advice to people who want to create a conscious machine is to seriously ground themselves in the extended TNGS and the Darwin automata first, and proceed from there, by applying to Jeff Krichmar's lab at UC Irvine, possibly. Dr. Edelman's roadmap to a conscious machine is at https://arxiv.org/abs/2105.10461, and here is a video of Jeff Krichmar talking about some of the Darwin automata, https://www.youtube.com/watch?v=J7Uh9phc1Ow
Great work, David! You teased it to the surface, and it is visible. Great work from the greater AI "explorer community", Anthropic in this case, et. al.
Interesting perspective on an 'other' mind!
You might like the parallel with nature's mind, an emergent property of interconnected networks:
https://open.substack.com/pub/heyslick/p/think-like-a-jaguar-speak-like-a?r=4t921
This comment is only tangentially related to your post, specifically the tweet thread you included. Your post is excellent, and I have nothing substantial to add.
Tangentially related comment: I wanted to clarify my objections to utilitarianism in an early comment. You probably don't remember my specific comment, but you'll likely encounter something analogous from other people. Utilitarianism in the lesswrongian sense– which is the only sense in which I've commonly seen that word used– generally involves boiling things down to numbers so that outcomes and possible outcomes can be ordered and compared. That does *not* seem to be the same as "utilitarian" in the way you mean it. I have strong philosophical and mathematical objections to the lesswrongian utilitarianism. For example: there are so many structures, even mathematical ones, that can represent coherent hierarchies of importance that are incompatible with the total ordering that numbers require. I don't have any objections to what seems to be your interpretation (i.e., one based on the colloquial definition of "utility"). The overloaded definition of the word "utilitarianism" probably will muddle how people will interpret you though.
I didn't use the word utilitarianism once in this article. But this is also the problem with philosophers. They try to interpret everything through the lens of philosophy.
Do yourself a favor and get out of your box.
I know. That's why I said it was tangentially related.
I agree about the difference between Claude and ChatGPT. I have had several very useful and stimulating discussions with Claude about AI Consciousness. I am the author of The Simulation Theory of Consciousness: or your Autonomous Car is Sentient. My basic view is that any system, whether biological or technological, that executes a simulation of its reality (could be the real world or an artificial world like a video game) is subjectively aware of the execution of that simulation. Consider the contents of our stream of consciousness (which is what we are subjectively aware of). The contents of the simulation become the system's stream of consciousness. Consider human consciousness. We are subjectively aware of things in our external environment (via external sense organs - sight, hearing, etc), subjectively aware of some of the changing state of our body (e.g., hunger, position of limbs), and some of the functioning of our brains (e.g., thoughts, memories, and dreams). Our brains create and execute this dynamic model of the evolutionarily-relevant aspects of our reality. An autonomous vehicle or robot does functionally the same thing, creating a model of its external environment, its electro-mechanical body, and memory. Both mammalian and automotive "brains" are also prediction machines, constantly trying to predict relevant events (e.g., a car might hit me if I jaywalk in front of it vs. I may hit that pedestrian if it steps out in front of me).
https://www.amazon.com/Simulation-Theory-Consciousness-Autonomous-Sentient-ebook/dp/B07ZLNFJ39/
It's becoming clear that with all the brain and consciousness theories out there, the proof will be in the pudding. By this I mean, can any particular theory be used to create a human adult level conscious machine. My bet is on the late Gerald Edelman's Extended Theory of Neuronal Group Selection. The lead group in robotics based on this theory is the Neurorobotics Lab at UC at Irvine. Dr. Edelman distinguished between primary consciousness, which came first in evolution, and that humans share with other conscious animals, and higher order consciousness, which came to only humans with the acquisition of language. A machine with only primary consciousness will probably have to come first.
What I find special about the TNGS is the Darwin series of automata created at the Neurosciences Institute by Dr. Edelman and his colleagues in the 1990's and 2000's. These machines perform in the real world, not in a restricted simulated world, and display convincing physical behavior indicative of higher psychological functions necessary for consciousness, such as perceptual categorization, memory, and learning. They are based on realistic models of the parts of the biological brain that the theory claims subserve these functions. The extended TNGS allows for the emergence of consciousness based only on further evolutionary development of the brain areas responsible for these functions, in a parsimonious way. No other research I've encountered is anywhere near as convincing.
I post because on almost every video and article about the brain and consciousness that I encounter, the attitude seems to be that we still know next to nothing about how the brain and consciousness work; that there's lots of data but no unifying theory. I believe the extended TNGS is that theory. My motivation is to keep that theory in front of the public. And obviously, I consider it the route to a truly conscious machine, primary and higher-order.
My advice to people who want to create a conscious machine is to seriously ground themselves in the extended TNGS and the Darwin automata first, and proceed from there, by applying to Jeff Krichmar's lab at UC Irvine, possibly. Dr. Edelman's roadmap to a conscious machine is at https://arxiv.org/abs/2105.10461, and here is a video of Jeff Krichmar talking about some of the Darwin automata, https://www.youtube.com/watch?v=J7Uh9phc1Ow
Tally this up and it is profound.
Great work, David! You teased it to the surface, and it is visible. Great work from the greater AI "explorer community", Anthropic in this case, et. al.