Video

fascinating stuff thanks all!

With this being presented as a salon, I think including a chat aspect to the conversation might open new territory to cover, thank you for sharing your time and work everyone, peace Great question to get clarity Alvaro, I really enjoyed the host's emphasis on communication in a concise manner in order to maintain fluid understanding of the unique problems you're engaged with here, peace

czcams.com/video/R6G1D2UQ3gg/video.htmlsi=Ne-ZjvbYuSf4V-Ih

thank you for a wonderful presentation and discussion

I think there is too much variation in mental disorders to categorize all of them as either brain disorders or not brain disorders, which has implications for public policy. One of the biggest obstacles to good public policies that respect civil rights (e.g., involuntary/voluntary medical treatment, housing/homelessness, culpability/innocence of persons for behaviors while living with mental disorders) is accepting that these are complex issues that require complex policies. One approach does not fit all conditions.

I think the issue is more one of ill-defined categories than a failure to find disorders in the brain. Of course we can find causes of disordered function in the brain. The problem is that the brain is a largely intractable, indivisible model of the world and itself. While mental disorders are simple models of culture that don't actually require brains. With sufficient technology and statistical modelling in centuries to come, we'll surprise ourselves with how well we can map the two. But as culture changes, so do the boundaries of our categories, and so we will learn to look for different patterns rather than discover the patterns we have been searching for

It's pretty enlightening. Thank you so much.

The top isn't in one part of the brain, it's the product of multiple parts working in concert, PFC, basal ganglia, cerebellum, etc... The same goes for the other levels of the hierarchy, on the input and output side.

So important as we move forward that disambiguation of language is an area that deserves serious attention and consideration, peace

Hello Maxwell, would you be happy with a description of what is within the Markov blanket as a "weakly asymmetrically connected system?" Can you recommend any grad/postgrad lectures on the subject? Is there anybody in the field interested in autistic spectrum behaviour disorders? My guess is, this might be systems getting stuck in some "free energy landscape traps", if you will. Thanks for you work Suav

I appreciate the access to really granular insight into something I have no connection to. It’s one of the promising features of the modern age. Thank you for your work and for sharing it.

From the FEP perspective, are the phenomena of conscious experience computations or are they non computational. like muscle, bone, or ligaments? Explanation?

The notion of a homunculus always seemed absurd to me. Of course, we are presented with a theater. Simply open your eyes and look around. There it is, playing in technical, surround sound, and smell-a-vision. It is a single impression at any moment of time. It is never, say, the semitransparent overlay of multiple scenes. The International Standards Open Systems Interconnect (OSI) reference model for computer to computer communication is a good metaphor for consciousness, and may apply more than metaphorically. The OSI model articulates seven levels of communication functions that allocate tasks like physical connection, error correction, message assembly, right up through presentation level to the top level called application. The application level then has the complete message and can act on it as it is individually designed (generative model) to do. Your computer screen is the presentation (6th level) of your computer's operating system. A biological system is the application in this sense, and consciousness is it presenting (as the presentation level) to itself. A theater? Yes. A homunculus? No. There has to be a top level. That is because the organism must act as a unit in response to its environment, i.e., all the other "things" out there. That is why perception only presents one interpretation of now. It does not present multiple possibilities because time (to survive) is limited. You cannot fight and flee simultaneously, nor can you take much time deciding. Interior Markov blankets can operate simultaneously. They generally do. But, their internal models are modified to support acting as a consistent whole organism. These modifications may be best explained through bioelectrics as articulated by Michael Levin. Edit: The homunculus notion is usually illustrated in the visual context. How does it work if you change that modality? Say, you feel something with your fingers. How would the little man inside your head replicate that. What would he touch? How could one possibly construct a Cartesian Theater in this case? How is homunculus universally applicable? Therefore, why does it merit consideration, at all?

This is a fascinating thought experiment. Personally, I think that if every neuron in the brain was decoupled then one would not have consciousness.

00:04 Jaan Aru presents the shallow brain hypothesis in the neuroscience and philosophy salon. 02:49 Most neuroscientists focus on the cortex and disregard the importance of subcortical structures like the Thalamus. 08:33 The shallow brain hypothesis challenges the dominant idea of cortical hierarchical processing. 11:28 Each cortical area sends output to subcortical structures 16:55 The brain has a hierarchical structure with modules and loops that can affect behavior at an early processing stage. 19:17 Modular architecture and updating less synapses is important in the brain. 24:17 The shallow brain hypothesis suggests that subcortical areas like the basal ganglia help with combining higher-order thinking and cognition. 26:43 The shallow brain hypothesis suggests that architectures with output from each cortical area could outperform deep architectures in tasks requiring compositionality and flexible combination. 31:48 Subcortical areas may have a cleaner and more primitive processing compared to the cortex. 33:55 The hypothesis suggests that the prefrontal cortex is a special place where everything comes together. 38:22 Loop theory supports working memory and attention 40:41 The shallow brain hypothesis suggests that the depth of the brain doesn't matter based on anatomical facts. 45:36 The amygdala plays a unique role in the shallow brain hypothesis. 47:49 The brain is better understood as 100,000 cortical agents trying to affect the subcortex. 52:27 The brain is not shallow, as it has complex hierarchical models and compositionality built on top. 54:31 The brain has shallow anatomical connections and deep architectures in Deep learning. 58:59 The human brain has a significantly higher number of cortical columns compared to animals. 1:01:17 Layer 5 PT cells project to the thalamus, modulating thalamic processing and driving higher order thalamic structures. 1:06:12 The brain has different processing modes and subcortical connections can handle novel stimuli. 1:08:48 Thalamus plays a crucial role in brain processes and is more intriguing than the LGN. 1:13:31 The speaker wraps up the discussion and thanks everyone for participating. Crafted by Merlin AI.

Great talk thanks. Are concepts such as: "Autonomous" and ""Autonomic", inherent and or emergent in this approach. I can appreciate that with the hierarchical and coarse graining considerations that a cognitive agent might have some potential for both. I'm curious about the value of come "cross cutting" capability ( possibly context aware ) in the policy selection that might be triggered Autonomically in the face of a threat and therefore interrupt a currently executing plan. Similarly but higher up in one's hierarchy of needs one might decide Autonomously to switch plans from the growing cascading wight of the underlying predictive coding errors coming from below.

what a crew !

The write only Markov blanket is more like a Schopenhauerian will, rather than a Cartesian rational agent.

Well, for what it's worth I think that the 'screen of cognition' is actually an affine separating hypersheet composed of an aggregation of hyperplanes generated by massive assemblies of small world networks applying differential and integral calculus, (classification and regression) rather like SVM. . .. . I put it to Dr. Hannah Fry a few years ago. She replied 'possible'. This fits well with the Dr.Eric Berne SA/TA transactional analysys psychological model . (PAC). . but that's just my view. I'm probably wrong.

So I came up with this.. Cross referencing between any different but correlated linguistic description produces contextually emergent clarifications that are the equivalent of what the Mandelbrot brot set would be to prompt engineering ...meaning the output of an operation becomes the input of another So to give a better example, Emergent Cross-Referencing Clarifications produce emergent contextual clarifications and it so far is but one side of the same coin like..When walking through a doorway, one side of the exit becomes the entrance to the same door, this door analogy is a direct example of the Mandelbrot set I came up with yesterday technically. The output of an operation becomes the input of another. The part above I specifically I had to repost from my Reddit .hope this recursion perspective helps! If you do use this for prompt engineering make sure to clarify and identify several types of recursion that may be present in the text. Don’t have to but it just goes to show what I’ve come up with can do that regardless .

Trying to see the forrest by looking at the leaves.

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 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 arxiv.org/abs/2105.10461

Man I miss this kind of atmosphere in academic discussion! Hope to see this kind of discussion live in the future

I find myself switching back and fourth in my support for the different views over the course of this discussion. Either I am not agnostic enough, or too agreeable, or the participants actually exchange really good arguments to support their views. I'd like to think its the latter, hihi. Thank you for reposting the meeting on youtube.

Subjective self refers to the ego in various vedic texts

nice discussion

Thanks for sharing! I think this is a useful perspective that clarifies at least that some conceptualizations of brain representations are actually compatible with enactivism (not the radical branch of enactivism)

These are really fantastic. Thanks, Luiz for putting them on, and Ann-Sophie, Antonella, and others for great discussion!

Is there a mailing list one can be on to know about upcoming NaPSs?

Very interesting! Will you present Mr Cisek's recording online this week? I hope the recording worked...

It's becoming clearer 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 primary consciousness will probably have to come first. The thing 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 arxiv.org/abs/2105.10461

A wonderful discussion. Thanks for sharing