We are a open-science Institute that is researching, teaching, communicating, and practicing applied Active Inference at www.activeinference.institute/
A very insightful video, thank you. It seems like some of the best young neuro scientists are being encouraged to use their knowledge of human brains to artificially re-create it in machines. Quite why their work has to be kept secret I don't know as if it's for the common good it should be openly discussed. Maybe their efforts will be used to control actions through words and for machines to run communities. This might look like humans are organising things themselves but who set the rules of the game and monitors the winners and losers?
Maybe the speakers are themselves living in a system that encourages them to evolve the system onto digital technology. We are born into the system and these two have been selected for their skills to take it to the next stage.
Would love to hear a discussion between Albarracin and journalist Sarah Kendzior about this topic and the latter's book "They Knew". But let's face it, that kind of crossover will probably never happen haha. Thanks for the talk
What frustrates me is that people think they have a solution, but that solution involves nano-tech and frequencies required for Web3 spatial computing that will disrupt the very life forms in nature you all seem to be concerned about protecting. Are we supposed to be happy with "successful" social impact climate change financial securities that coat every forest in nano-dust and monitor it and the villages of the Amazon from space in order to earn a good citizen badge in the leaderboard battle of carbon reduction? It's so gross. You all really have NO idea what you are actually building. Tyranny of narrow-minded experts. None of this is new John. Are we supposed to pretend that we don't understand that your collaborator spend 10 years at Orange? C'mon. We're not all dummies here.
You are using rose-colored glasses to justify your biocybernetic social program. Has it occurred to you maybe we don't want to live in an outside in robot so impact investors can get their digital evidence to legitimize gross profit taking on poverty management and cut rate environmental remediation ? Let me guess...one of the social goods we'll be pursuing is investments in early childhood education and pre-natal care. I'm telling you I've been researching this for over a decade. Are you aware that the plan by the large hedge funds is to bring pay for success deals to scale and bundle low-income preschoolers in headstart into derivatives securities to trade on global markets? Yeah John, that was the plan in 2012 BEFORE your three papers. I'd love to know who's funding you, because they're probably also funding Michael Mateas's work, too. Is it INET? If not I'm guess you'll hit up Soros soon. This is NOT ok. And just so you know I used to be liberal and left and now I'm not anything, because the people I thought actually cared about social justice are all either idiots or in league with the billionaire class. web.archive.org/web/20191113172246/www.payforsuccess.org/sites/default/files/resource-files/20120403_KauffmanReadyNationPKSEPresentation_0.pdf wrenchinthegears.com/wp-content/uploads/2018/03/kauffman-readynation-sib-2.jpg
I can't believe I watched the whole thing. His system is a colossal waste of time and human capacity. Maybe they are planning on having a large, low IQ population to manage. Life doesn't work this way. Models are static, not creatively present. Most of the stuff he thinks of as collective mind work already is built into the human system. In a healthy, natural system, we tune in to one another and all of that is going on in a wholistic way. Systems like the Catholic Church and others exploited this capacity by creating a standard model and regular gatherings to create a culture of compliance. Its a powerful thing when minds group together for an altruistic motive. Less powerful when done with lower intentions.
Phenomenal! Lays a rigorous foundation for Active Inference! Well done! Takes Active Inference to another level! And makes it implementable and scalable - and auditable!
the Tensorial relationship between InTentional Focus and Mechanistic AtTention Intellidoscope equation including XZY+Q Coordinates Let 'T' Stand for Tensorial Let 'I' represent Intentional and 'F' be Focus Let 'M' represent Mechanistic and 'A' Attention. •X( zi Fqt(mA ) Z𝐈 ( aM)TQf zi )Y•
Personal preference for music other than playing or listening to live music is a curated playlist of 5000 tracks and soundbites on random with 9 sec fade crossovers.
We can change evolution or at least tweak it and I think that's what we're doing with meaning of life, which to me looks like being grateful to be alive. We've tweaked evolution by wearing glasses why not by realizing our true life has truth, meaning and a purpose.❤🙏❤️
Leveraging Inference for Efficient Data Representation. Optimal Compression of Data and Efficient Representation. Intellidoscope compresses complex data into compact forms without losing relevant information through its unique algebraic calculus and syntactic framing. Dynamic Inference: It leverages inference to dynamically deduce relationships, optimizing data compression. The standard syntactic framing of Intellidoscope •(()())• can be reduced further when represented in ‘Short Form’ •((). Optimal Representation of Relationships Simultaneous Representation: Unlike many traditional methods, Intellidoscope simultaneously represents elements and their relationships, in time providing a more intuitive view of the data. Comparison with Other Methodologies INTELLIDOSCOPE SHORT FORM •( c b(A )C SYMBOLIC AGGREGATE APPROXIMATION (SAX) [a, b, c] Transforms continuous data into discrete symbols, but may not capture complex relationships. SP THEORY OF INTELLIGENCE Pattern 1: [A, B, C] Pattern 2: [A, X, C] Alignment: [A, (B|X), C] Emphasizes pattern matching and unification, which can be complex and less intuitive. HIERARCHICAL COCLUSTERING Cluster 1: {a, b, c} Cluster 2: {d, e, f} Hierarchy: {{a, b, c}, {d, e, f}} Encodes hierarchical relationships, which might be too rigid for dynamic inference. NEURAL NETWORKS Input: [x_1, x_2] → Weights: [w_1, w_2] → Output: y Compresses patterns but often lacks interpretability and requires extensive training. EXPENSE OF DISCRETIZING SPACETIME Discretizing spacetime is computationally intensive and may lead to unnecessary energy consumption. APPROACH: Leveraging inference and equivariance, as demonstrated by Intellidoscope, particularly by utilizing Intellidoscope Short Form, offers a more economically responsible way to model complex systems. This approach can optimize computational processes and reduce unnecessary energy expenditure while ensuring that symbolic representations remain consistent under transformations. This is particularly useful for maintaining the integrity of relationships and structures. EXPLORING QUANTUM GRAVITY AND DISCRETE SPACETIME WITH INTELLIDOSCOPE. Continuous vs. Discrete Spacetime: In classical physics, particularly in Einstein’s theory of general relativity, spacetime is modeled as a smooth, continuous manifold. This allows for the seamless description of gravitational fields and the curvature of spacetime due to mass and energy. Smooth Manifolds: Continuous models treat spacetime as a smooth fabric, where points are infinitely close to each other, allowing for differential calculus and smooth transformations. Discrete Spacetime: Some theories in quantum gravity, such as loop quantum gravity, propose that spacetime might have a discrete structure at the Planck scale. However, this idea can be seen as mistaken because it leads to exponential use of computational energy and overlooks the advantages of inference and compressed symbolic language. Computational Models: While discrete models can be useful in simulations, they may not accurately reflect the true nature of spacetime and can be computationally expensive. Intellidoscope’s Perspective: Instead of explicitly discretizing spacetime, we can use inference to dynamically represent relationships and structures. This approach can maintain the continuity of spacetime while optimizing computational efficiency. Equivariance ensures that representations remain consistent under transformations, supporting the idea of a continuous spacetime where relationships and structures are preserved. Cross-Categorical Modeling Modeling with Intellidoscope reveals a cross-categorical approach: The quantum realm appears to be continuous, aligning with its smooth, probabilistic nature. Gravity, on the other hand, may exhibit discrete characteristics. Distinguishing Point Space from Discrete Units POINT SPACE: In quantum field theory, point space refers to the continuous field where each point in space can be associated with a quantum state. DISCRETE UNITS: The idea that spacetime is composed of discrete building blocks is seen as an error. Instead, we should consider the concept of universal discrete materiality within the continuous quantum field of point space. SUBJECTIVE POINT SPACE: Points within this point space can be considered subjective, as they are points that we consciously choose to focus on. This subjective consideration allows for intelligent and efficient modeling. INTELLIDOSCOPE EQUATIONS Spacetime Continuum: •X(s zc qt() ZC ()TQ zc S)Y• Point Space over Discrete Units: •X(s zP q(δu )Z( U∆)Q zp S)Y• Quantum Gravity: •X( z q(g )Z( G)Q z )Y•
In 1:12:00, we see this - B_context[:,:,0] = np.eye(len(context_names)) It’s telling us that if the action is ‘Start’ then if we had ‘Left-Better’ then that doesn’t change and we are left with ‘Left-Better’ and if we had ‘Right-Better’ then that doesn’t change and we are left with ‘Right-Better’. However this should be true also for the other choices: 'Hint', 'Left Arm' and 'Right Arm'. So I think we should add B_context[:,:,1] = np.eye(len(context_names)) B_context[:,:,2] = np.eye(len(context_names)) B_context[:,:,3] = np.eye(len(context_names)) No?
@RichBlundell you’ve tapped into a core (the barrel) of Karl’s journey. Watch the energy build, marry your momentum to that of the direction of travel, ride, then decide when you want to give it another go. Although Karl and Kai Lenny tow in different conditions, I somehow think of both as immense creatives at what they do.
totally. And now I want more conversation in our culture about what all this means for our conceptions about who we are and how to live more aligned with the ontology (reality) that these insights can (and should) have. I get the conceptual and theoretical aspects, now what about the axiological? In other words, if these dynamics hold up to scientific scrutiny, or at least consensus (and I'm convinced they do both), how can they change our "collective" mind? Who out there on the interweb is communicating them into a rational and legitimate cultural-level conversation so they can be integrated into civilization?
I am a surfer. I can't help but get the sense of surfing while comprehending all this. It's because Karl Friston is a surfer too. He is expertly surfing the gigantic waves of a confused sea. He surfs the deepest continuities of math, physics, deep time (with the prebiotic soup), computational neurophenomenology, and neuroscience. The paradigmatic impact of what he's saying can't be overstated.
This is very good lecture and I learn't a great deal from it. However, Ben keeps mentioning a 'document' which I assume has a full bibliography. I've already downloaded the slide deck of the lecture but can't find the 'document'. Am I missing something obvious? Can some kind soul give me the link?
Could be it this link? coda.io/@active-inference-institute/active-inference-social-science-aii-2023/course-syllabus-2 Or let us know. Thank you for watching.
Spin Mass Artistic representation (⬇️ ⬆️🎗️⬇️ ⬆️🎗️⬇️ ⬆️) Intellidoscope •X(s z q(m )Z( M)Q z S)Y• Dimensionality of Intellidoscope d7•d6(d5(d1 d2)d5(d3 D4)d5)d6•d7 Beyond D4 Higher Order Quantum Sensing •X h(s z q( o)Z(O )Q z S)H Y• Ntanglement N•X( z q()Z()Q z )Y•N X is to Y as Y is to X Redshift/Blueshift and Spin States What next cognitive light cones?¿😎
Cool, an actual use case for violin plots! I've heard them badmouthed by some data visualizers, but I find them aesthetically information-packed and pleasant to contemplate.
👋😁 Thanks for putting these videos up I find the concepts around Active Inference inspiring. INTELLIDOSCOPE EQUATIONS: Agentic Isomorphic System •X(s zi q(A ) Z𝐈 ( a)Q zi S)Y• Möbius Strip of a Torus •X(s z qt(m )Z( M)TQ z S)Y• While a Möbius strip and a torus are fundamentally different in their topological properties, you can create a Möbius-like path on a torus by tracing a line in a specific manner. For example, if you trace a line that loops around the torus in such a way that it covers both sides of the surface, it can resemble the continuous nature of a Möbius strip. Old MacDonald... •Xe ( zi q( Eo) Z𝐈 (Oe )Q zi ) eY• EIEIO • Energy: Symbolizes energy over a system transform • Integration: Reflects the continuous realignment of system coherence. • Optimization: Indicates minimizing resistance. Self-Organising Principle •X(s Pz q(o )Z( O)Q zp S)Y• The Ability of Self-Reflection or Systemic Agenetic Referentiality. •X(s z Rq(A )Z( a)QR z S)Y• the Sensor the Reflection and the Mirror •X(s z Rq(m )Z( M)Qr z S)Y• Homeomorph beyond the Isomorphic Continuity •X h( zic q() Z𝐈C ()Q zic )H Y• • The Isomorphic continuity of the fifth dimension, preserves the structural integrity of Spacetime. • The Homeomorphism extends beyond , into the sixth dimension. This dimension can be thought of as an energy environment beyond mass and motion.
22:00 Entropy in classical information theory is better thought of as "missing information" instead of "information". In fact that was the term Shannon originally used. In those terms, it makes sense to say that when all the components are independent and entropy is maximal. When you observe events from this system they will have a higher "missing information" burden (the information needed to encode the event, which you did not possess prior to observing the event) since no system component carries information about the other components.
I suffered pre-verbal childhood trauma; this feels like a Black Hole in my mental/emotional experience. A consistent pull downward, largely subconscious. We can’t see gravity but we can see its effects. The effects of my black hole in my life are obvious, and pills cannot change my experience of said black hole. I am finding this a useful model, a good analogy. Edit: 31:23 is my lived experience. My ex once said “many people fall down, you never got back up”. Gravity was too strong for me, especially when the ‘load’ of a mortgage and 3 kids was added to me. Getting ‘high’ (pot) temporarily gets me out of the gravity well; with very clear perspectives on the initial high. But I have a hard time functioning on it. I really find this a useful model. Thank you for helping me understand my experience.
The Interplay of Wave Functions within Neural Synchronization. N•X(s zi Fqw() Z𝐈 ()WQf zi S)Y•N Bifurcation waveforms appear to underpins memory processes. The fine-tuning of biases that leverages the shape of neural systems can be modeled by a tenseors or transformers as both biological and artificial, memory is continuously reframed and realigned within an ever-evolving quantum computational framework. This all has some interesting implications for our experience of time. Each fine-tuning of biases generates a quantum shift that represents a morphism in our limited perspective of objectively held memory. I wonder about the rate of morphogenic change? 🤔 Anyway. Storage and Retrieval Function over Time of Objective Memory •X(s z RFqt(mo) Z Leveraging the isomorphic structure of Intellidoscope allows for a form of symmetry that can be exploited, compressing information further, optimising memory usage. By understanding one part of the structure, the other part can be inferred, almost halving equation size. Just as there is standard ZX Calculus, therefore we call this 'XZ Intellidoscope'. Subjective Brain Waves and the Morphogenic Material of Mental Gravitas: •X(s z qwb(gm )Z( MG)BWQ z S)Y• further investigate regarding this equation, as 'surprisingly' the relationship between the brain and mind in terms of objective and subjective has gone through a inverse transform put simply it is somewhat unconventional to think of the brain existing within the mind. That said, there's a realization that we are dealing with a higher dimensional shape. Hyper-MultiWay System •X h(s z qw(m )Z( M)WQ z S)H Y• A Hyper-MultiWay System extends the concept of multiway systems into the realm of hypergraphs. It represents all possible states and transitions within a system, capturing the dynamic nature of complex interactions. In a Hyper-MultiWay System: - Hyperedges connect multiple nodes, representing complex, multi-entity interactions. - Multiway Transitions capture all possible states and transitions. Hyper-MultiGraph •X h( z q(gm )Z( MG)Q z )H Y• A Hyper-MultiGraph combines the concepts of multigraphs and hypergraphs, allows for multiple hyperedges connecting the same sets of nodes, representing complex, multi-entity interactions with multiple layers of relationships. In a Hyper-MultiGraph: - Hyperedges connect any number of nodes, representing complex relationships. - Multiple Hyperedges can connect the same sets of nodes, capturing the richness of interactions between entities.
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 arxiv.org/abs/2105.10461
