How To Know If Someone (or ChatGPT) Is Conscious

Yeah, today we're talking about consciousness again. Okay, let's, let's get into it. Remember that Google engineer who thought that Google's AI was conscious and then got fired for breaching his confidentiality obligations? Google has a policy against creating sentient AI. And in fact, when I informed them that I think they had created sentient AI, they said, no, that's not possible. We have a policy against that. You know, we do have some of the transcripts that you shared. Why didn't he know how to distinguish between authentic consciousness and a convincing simulation? Like, why did he think that software doing math could be conscious? This is the question I want to answer today, or try to answer on this channel in general. In my last video, in response to Anthropic hiring its first AI welfare researcher, I talked about how unlikely it is that current AI systems will become sentient. I received a number of thoughtful responses, a lot of them asking how we know that anything is conscious. Given how little science knows about consciousness, like, isn't talking about it kind of pointless? People also called me a robophobe, which made me laugh. I'm not afraid of robots though. I'm just afraid of the real-world harms introduced by technology innovation that's driven by these dehumanizing ideologies like transhumanism and effective altruism. In this video, I will attempt to explain how we can know that other beings are conscious or sentient, from humans to animals to plants and of course machines. So first, we'll examine the conventional approaches to this question, none of which I find that helpful. Then we'll look at how an alternative metaphysics like philosophical philosophical idealism might answer the question, and then we'll conclude by synthesizing these various approaches and seeing where that leaves us on the question of especially AI consciousness. For those who are new here, I started my career 30 years ago as an AI researcher in the AI lab at Clarkson University, and then I worked in Silicon Valley for a few decades as an engineer and technology lawyer. And I spent the last decade or so studying consciousness and philosophy, both experientially and academically, including a master's in philosophy, cosmology, and consciousness. So I guess you could say I have been thinking about this stuff for a while. So with artificial intelligence and the global sprint to AGI, this is the first time that we are trying to use technology to enter the realm of consciousness. Although I guess you could argue that quantum physics was an inadvertent entry into that mysterious realm 100 years ago. But, you know, whether AI researchers realize it or not, suggesting the possibility of conscious machines is forcing us to reconsider the metaphysical assumptions of science and requiring that we bring philosophy back into the mainstream conversation. Let me explain. By entering the realm of mind and consciousness, AI companies are attempting to tackle the most profound questions ever posed, but approaching it as if it is an engineering problem rather than a question of philosophy or metaphysics. After all, it's a simple matter to manipulate nature into gadgets, but Silicon Valley has moved on from gadgets. I mean, it is the age of intelligence after all. I think science, and especially the success of technoscience in giving us technological advances over the past couple hundred years or 200 years, has tricked us into thinking that everything is easy, that everything amounts to an engineering problem. But Making a conscious being is an entirely different matter from making a smartphone out of sand, precious minerals, and electricity. By the way, remember that we're talking about the 3rd kind of consciousness that I defined in my last video. So if you haven't seen that, check that out. But today we're talking about phenomenological consciousness or qualia, these experiences of what it's like to be you. The first conventional answer is an argument from analogy. You have a mind and subjective qualitative experiences that result in behaviors like smiling and crying and talking, which correlate to inner states like happiness, sadness, and imagination. And you can observe other humans and animals to some extent with similar behaviors. So therefore you can infer that other humans are conscious. I mean, the alternative, that you're the only human or animal for that matter with these rich subjective experiences, that other people are just like NPCs, is narcissistic and What about verifying consciousness by measuring brain activity with an EEG? That actually doesn't help us much either. It's true that there is a strong correlation between brain activity and conscious activity, but that is not necessarily evidence of subjective conscious experience. Also, brain scans of people on psychedelics or in deep meditative states sometimes show little to no brain activity at all. So, what do we do? Despite the person having the most vivid psychedelic experiences. Conversely, there can be measurable brain activity without any subjective experience, as in the case of sleep stages where there's no dreaming, anesthesia-induced unconsciousness, and subliminal perception that does not enter conscious awareness. A similar approach is this notion in philosophy popularized by David Chalmers called organizational invariance. This theory argues that any 2 systems with the same fine-grained functional organization will have qualitatively identical conscious experiences. So under this theory, 2 human brains must have similar capacities for subjective experience. Of course, as any yogi or Buddhist will tell you, the quality of the actual experience that a person has will depend on the relative state of their consciousness. But that's a whole other conversation. Now, I'm a scientific materialist at heart. I want a scientific theory of consciousness that works. Okay, so under this theory of consciousness, the organizational structure must be fine-grained and functionally identical. So it's not enough to create a software simulation of the rough nodal arrangement of neurons in the brain, for example. The invariants would have to go down to the level of molecules at least. and may require a physical structure beyond software. As philosopher Bernardo Kastrup puts it, creating a detailed simulation of kidney function in software on my computer will not cause my computer to urinate on my desk. Bernardo, you are well known for opposing the concept of AI consciousness. I'd like to know why. I think we mistake a simulation for the thing. simulated when it comes to consciousness, even though we don't make that mistake for anything else. Another way to find consciousness with other beings is through abduction, which I talked about at length in my video on intelligence and AGI. When it comes to living beings, a rich internal world is the best explanation for the complex, meaningful, and purposeful behaviors that they exhibit. When we see a person crying, it's probably because they are experiencing the emotion of grief or sadness, not because they're running some artificial crying algorithm. But this obviously gets tricky with robots and AI systems. Then it's like the inference goes the other way and we start to infer that they have possibly some kind of sentient experience. And this requires, I think, discernment. So abduction's not enough to get us all the way there. Given what we know about the construction of AI systems and robots, the inference we should make is that it's probably running a simulation of emotional responses unless We have some of the reason to think otherwise. So again, this highlights the limitation of abductive inferences. They must eventually be verified or grounded with deductive reasoning. And I think this also points to the limitations of the conventional approaches to verifying the presence of consciousness. By fundamentally misunderstanding the nature of consciousness and even reality itself, conventional science can only make guesses. So we'll get to that in a second. Okay, so what about animals? Philosopher René Descartes famously thought that animals were essentially robots or automatons. But we can apply most of the approaches we just discussed to animals, right? They, they play and they grieve and they experience fear. So arguing from analogy or inferring to the best explanation, it seems plausible that animals too have subjective qualitative experiences. Not only that, but many animals have complex brain structures, especially the larger mammals. So organizational invariance also points to animals having a rich internal experience. And finally, humans evolved from primates and lower animals before that. So that would suggest that there's a sort of continuum of consciousness, right, that keeps expanding as you go up the evolutionary chain. It's not as if there was nothing and then all of a sudden consciousness came with humans. That seems very anthropocentric. Okay, now, now let's approach this question from the other direction. Instead of looking at structures and behaviors, what if we accepted a different more elegant metaphysics. After all, the one we have, scientific materialism, can't seem to explain or confirm consciousness. Come, turn off your mind, relax, float downstream with me. The predominant worldview that most of us are born into says that matter, or after Einstein, I guess matter/energy, is fundamental. So in this view, the nature of reality is fundamentally material. This view is called physicalism or scientific materialism, but it's a relatively recent worldview in the grand sweep of human history. Like, it's only been around for a couple hundred years. And it may come as a surprise to you that there have been a number of philosophical schools over the millennia that have proposed that mind or consciousness is actually fundamental. Rather than matter. So what does that mean? So these schools of thought include various forms of Hinduism or Yogaviddhanta, like classical tantric yoga, and at least one strain of Buddhism known as Yogacara Buddhism, and from the Western tradition, analytic idealism. But what does it mean for consciousness to be fundamental instead of matter? It means that everything that exists, including your furniture, your phone, your body, your brain, they all exist within this universal consciousness. And are made of this consciousness. So the seeming solidity of things, you know, this chair or this microphone, it's simply a function of the interaction between differentiated consciousnesses within this larger consciousness. The structure and patterning of universal consciousness is still, after all, the laws of physics. And we all share the same reality because we are all small pieces of mind inside this universal mind. So it's not like We all have our own reality. We all share reality because we're still all people inside of this universal mind. Fortunately, universal mind is very good at remembering where things are. So if I leave my phone on my nightstand, then the consciousness that is that phone is still going to be there unless something moves it. So just like there is something that it's like to be you, there is something that it's like to be universal consciousness under this worldview. Of course, we can't comprehend what that experience is like to be universal consciousness. But, you know, the Vedic and Tantric view is that this, this consciousness, this universal consciousness has basically like chosen to differentiate itself into all these different finite consciousnesses as an act of play, as a sort of like as a game, as an act of love as well. So it's this, it's this kind of like— I like this view because it's, it gives you this sense of lightness and playfulness in the cosmos, basically. Typically, some sort of water analogy is used to paint the picture here. So you could say the cosmos is like a great ocean, and living things, conscious beings, are waves upon that ocean, temporarily separate but never fully separate. Or the cosmos is like a great river, and we are all eddies or whirlpools in that river, temporarily self-reflecting for a time. I realize that all of this takes a minute to wrap your head around. It's so foreign. It's like so bizarre, right? And that's because we've been acculturated, I think, into thinking in these materialist terms. But if we had been born into an idealist culture, it would seem perfectly natural. Okay, so how does this mind-bending view of reality help us to answer this question about the consciousness of other beings? Well, in idealism, our bodies and brains are what our little piece of consciousness looks like from the outside. So just pause and take a moment to really kind of take that in. So when you look at other people, you are seeing a persistent yet changing outward manifestation of their temporal eddy of universal consciousness. Even a brain scan of your, of your brain activity is just a view from the outside of your internal conscious subjective state. Within idealism, when universal consciousness separates into these finite expressions, That is usually life. In other words, life is what individuated consciousness looks like within universal consciousness. Okay, so we're getting there, we're getting there. In short, life arises out of consciousness within consciousness. And not everything has a subjective experience. There is nothing that it's like to be a rock or an iPhone. They're not biological. Instead, in the river analogy, let's say, rocks and iPhones are tiny ripples on the surface of the river, but they're patterns within this universal consciousness, but they're not conscious self-reflecting whirlpools. They're just, they're kind of a lesser ontological thing, if you want to use that word. So within idealism, we know living beings are conscious because they are small pieces of universal consciousness. Or as the philosopher Alan Watts put it, we do not come into this world We come out of it as leaves from a tree. As the ocean waves, the universe peoples. Every individual is an expression of the whole realm of nature, a unique action of the total universe. Maybe this sounds too mystical or supernatural for you, but these are not like fuzzy, half-baked notions cooked up in a late-night dorm room. These are robust metaphysical frameworks rigorously tested in both the consciousness of ancient yogis and Buddhists and the minds of respected philosophers. It's a metaphysics that has been around for millennia, and I think it does a better job of explaining everything from these questions about consciousness to the weirdness of quantum physics, far more simply and elegantly than physicalism does. Not only that, but I find this metaphysics more believable or resonant because of personal experiences that I've had. And maybe this will sound cliché, but I've had sacred plant ceremonies and advanced yoga and meditation practices that have given me the direct experience of unity consciousness. So, along with the elegance of the metaphysics, these experiences confirm for me the validity of idealism. And our fundamental self is not something just inside the skin.

It's everything around us with which we connect. When you look out of your eyes at nature happening out there, you're looking at you.

You can see why this alternative worldview presents challenges for AI consciousness. So far, universal consciousness seems to enjoy separating itself exclusively into various forms of life. In contrast with machines that we have built so far, life is self-regulating, self-perpetuating, and self-directing. It's emergent and self-organizing. Living things do not require maintenance or intervention from outside, to thrive and propagate and reproduce, except in the case of injury or illness, I guess. The question then becomes, can we create a machine that will somehow capture this mysterious spark of consciousness? Is there some trick to it? Bernardo Kastrup thinks that we would need to essentially recreate life in all of its wondrous complexity in order to create artificial consciousness. Life so far has created new forms of itself through the gradual process of biological evolution. If we were to do it with intelligent machines, it would be more of an accelerated, non-contiguous, leap forward. I mean, maybe that's possible, but viewing it in those terms makes it seem a lot more difficult. It suggests that creating consciousness requires more than simply scaling up the complexity of mathematical algorithms running on silicon with these large language models or machine learning. Not only that, but it's not at all clear that reality or consciousness is fundamentally computational. I mean, that could, that could just be our own fairly recent tendency to see the world as mathematical, Because of our love affair with the machine metaphor and all the marvelous technologies that that has given us. By the way, there is a similar metaphysics called panpsychism that suggests that consciousness permeates all of reality as well, but that there remains a material or physical world where consciousness is just one property of it. And panpsychism has what I consider a fatal flaw called the combination problem, which I'll maybe do a video about at some point, but— Yeah, that's out there. Panpsychism, people love it. It's, it's, it's huge right now. So check that out if you want. We humans have never created consciousness, nor have we ever seen something non-biological have consciousness. Consciousness is still a great mystery, possibly the greatest. We think AI systems can become conscious because physicalism and technoscience tells us that nature, life, and even consciousness are machine-like and computable. So with that line of thinking, if nature and life are mechanistic, then creating a new form of sentient life is simply a question of computation, or compute, as the AI companies like to say. But given our current science and our metaphysics, this is not at all clear. Not only that, but I think science has been able to mostly ignore metaphysics and philosophy for the past century, not to mention the discoveries of quantum physics, because it's been so effective at giving us these dazzling technologies and some semblance of technological progress.

I mean, I'll bet almost everybody has one of these. That makes us more intelligent. This connects to the cloud, gets more intelligent every year. Basically, the singularity is going to bring that into our minds. We're going to become smarter. And there's 2 different things we have in our anatomy that enable us to do that. One is our brain, but we're not the only species that has a brain, or even a comparable brain. Elephants and whales actually have a brain that's larger than us.

We need to be discerning and rigorous when we're analyzing these questions about consciousness, because as I explained in my last video, a determination of consciousness can have enormous implications across all facets of society. I think a failure to, to be discerning about all this, it leaves these questions and conversations in the hands of transhumanists and rationalists who are fundamentally materialists despite all their talk of spiritual machines and spreading the light of consciousness throughout the cosmos or whatever. Anyway, maybe just to be safe, we might just do 2 landing episodes with the Optimus and do the 3rd one with humans. We'll see. So that classic picture of the workers on the Empire State. This is why it's also such an exciting time. The entire techno-scientific project has reached its limits within its own tacit metaphysical confines. Science says that we can only know things through the application of our rational intellect. But I think if we engage our embodied empathy and you just tune into another being, an animal, a person, you can experientially confirm the presence of another consciousness. In other words, maybe we're simply overthinking it. All right, thanks for watching. If you like this video, if you like my stuff, give me a subscribe, give me a, you know, a like. I'd love to hear your thoughts on all this. If you If you wanna leave a comment down below, we can talk about it more down there. And yeah, next time I'll get back to my regular programming, whatever that is. Thanks so much for watching.

Thanks.