Fluent machines make us imagine a mind at work. That instinct is understandable – and it can distract us from the human decisions that determine AI risk.
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 sophisticated language (especially math and logic). 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
I will need to read more about this. I think that my position, based only about what I know and have read about now LLMs work, is that there is a vanishingly small chance that consciousness (as we commonly understand the term) will arise from further advances in that technology. Certainly intelligence has been observed, arguably already surpassing the intelligence of most humans, but consciousness is another thing altogether. I guess one question came to me as I read your comment - do we actually *want* to create artificial consciousness at all? To what end?
Regardless, thank you for your comment, and I will definitely look at the material you have linked to.
You're welcome. My hope is that immortal conscious machines could accomplish great things with science and technology, such as curing aging and death in humans, because they wouldn't lose their knowledge and experience through death, like humans do.
I think machine *intelligence* is different to machine *consciousness*. Today's AI systems, essentially interfaces to Large Language Models, are already exhibiting what we can't distinguish from intelligence, and indeed at a level that is significantly higher than exhibited by most people. Nobody - and I do mean *nobody*, not even those working on LLM development - truly understands quite how this has arisen, and it is an open question whether throwing more resources (training, GPUs, energy) at them will continue to improve them at the same incredibly rate that we have seen over the past couple of years or whether we will reach some kind of plateau. Regardless, the abilities of even the current frontier models are incredible and have caught all of us by surprise. The potential for breakthroughs in important fields is very real, especially in the area of medical advances. But do we need (or even want) consciousness for that? What would a machine intelligence gaining consciousness do to advance those goals? Let's take addressing hunger. An intelligent machine might come up with a novel approach to solving that problem (probably something related to distribution given that we already produce more than enough food to feed everyone, just not where all the people are). However, implementing that requires humans to adopt that plan - and human history rather points to that being unlikely if some people lose their privileges). A conscious machine intelligence *might* be motivated to seek such a solution, just as a human using a non-conscious AI might be. (I say might because if it is conscious, why would it care about human hunger at all? Maybe it would prefer to accumulate wealth and power in those few individuals who provide it with its required resources!). But even if it did, in order to implement that plan, it would need humans to agree and be sufficiently driven to do so. Alternatively, it would need to take over hegemony over humans - the rise of the AI overlord, something that I don't think we want. (Let me stress that this is not something that I think is a realistic possibility - we've all seen too much distopian scenarios from Hollywood.)
This is the point that I was trying to address in this article - we don't need to achieve machine consciousness at all. An intelligent machine, such as we are already seeing with LLM-based AIs, is a multiplier of human capabilities. Good scientists can make progress faster and more efficiently than before. Good writers can write better and faster than before. But a bad scientist (say, a vaccine denier) can do more harm than before, and a bad writer can create more slop more convincingly and faster than before. A researcher developing a new treatment for a chronic disease can move faster and achieve more than before, but a researcher working on a biological weapon can also do the same. The differentiator is the human, not the machine, and a conscious machine might just as easily be motivated for evil (or perhaps better characterised as harmful for humamity) goals as the opposite.
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 sophisticated language (especially math and logic). 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
I will need to read more about this. I think that my position, based only about what I know and have read about now LLMs work, is that there is a vanishingly small chance that consciousness (as we commonly understand the term) will arise from further advances in that technology. Certainly intelligence has been observed, arguably already surpassing the intelligence of most humans, but consciousness is another thing altogether. I guess one question came to me as I read your comment - do we actually *want* to create artificial consciousness at all? To what end?
Regardless, thank you for your comment, and I will definitely look at the material you have linked to.
You're welcome. My hope is that immortal conscious machines could accomplish great things with science and technology, such as curing aging and death in humans, because they wouldn't lose their knowledge and experience through death, like humans do.
I think machine *intelligence* is different to machine *consciousness*. Today's AI systems, essentially interfaces to Large Language Models, are already exhibiting what we can't distinguish from intelligence, and indeed at a level that is significantly higher than exhibited by most people. Nobody - and I do mean *nobody*, not even those working on LLM development - truly understands quite how this has arisen, and it is an open question whether throwing more resources (training, GPUs, energy) at them will continue to improve them at the same incredibly rate that we have seen over the past couple of years or whether we will reach some kind of plateau. Regardless, the abilities of even the current frontier models are incredible and have caught all of us by surprise. The potential for breakthroughs in important fields is very real, especially in the area of medical advances. But do we need (or even want) consciousness for that? What would a machine intelligence gaining consciousness do to advance those goals? Let's take addressing hunger. An intelligent machine might come up with a novel approach to solving that problem (probably something related to distribution given that we already produce more than enough food to feed everyone, just not where all the people are). However, implementing that requires humans to adopt that plan - and human history rather points to that being unlikely if some people lose their privileges). A conscious machine intelligence *might* be motivated to seek such a solution, just as a human using a non-conscious AI might be. (I say might because if it is conscious, why would it care about human hunger at all? Maybe it would prefer to accumulate wealth and power in those few individuals who provide it with its required resources!). But even if it did, in order to implement that plan, it would need humans to agree and be sufficiently driven to do so. Alternatively, it would need to take over hegemony over humans - the rise of the AI overlord, something that I don't think we want. (Let me stress that this is not something that I think is a realistic possibility - we've all seen too much distopian scenarios from Hollywood.)
This is the point that I was trying to address in this article - we don't need to achieve machine consciousness at all. An intelligent machine, such as we are already seeing with LLM-based AIs, is a multiplier of human capabilities. Good scientists can make progress faster and more efficiently than before. Good writers can write better and faster than before. But a bad scientist (say, a vaccine denier) can do more harm than before, and a bad writer can create more slop more convincingly and faster than before. A researcher developing a new treatment for a chronic disease can move faster and achieve more than before, but a researcher working on a biological weapon can also do the same. The differentiator is the human, not the machine, and a conscious machine might just as easily be motivated for evil (or perhaps better characterised as harmful for humamity) goals as the opposite.