Видео

Don't understand why this isn't more popular.

Santafe nonsense

Beautiful presentation. I've been so fascinated with this concept for so long and after listening to and reading many experts' thoughts on it, one recurrence that always pops up is: predictability/certainty. Predictability seems to be a function of parameters of a lower-level space operating at the large distances from an "edge of chaos" parameter value - it creates a very predictable structure within a specific parameter space (e.g. AlphaGo operating in a predictable manner at room temperature). It isn't until a parameter approaches a value that the system's predictability begins to decrease (e.g. a heated room causing chaotic behavior on the level of transistors in AlphaGo, which ripple into the higher emergent spaces). Tim Palmer writes an excellent book on this topic ("The Primacy of Doubt") that connects with what David mentioned about how emergence is "bounded" by the spaces beneath it and will begin to break down at certain parameter values. I'm trying to self-study higher maths to guide my understanding of how emergent systems may arise at the chaotic boundary between phase transitions of these emergent levels, but I have to admit that sometimes the abstraction of this all becomes extremely difficult. Thank you for this presentation.

Fascinating.. David is really a visionary. He guides and combines this field's ideas and "axiomatizes them" - he makes sense from the mess of great ideas. I heard multiple complexity scientists talk about what is complexity science and usually they seem to start explaining its usage, combining with some "buzzwords" such as emergence - but David put it the best. When this field will have increasing importance and reputation (outside physics) and other fields/industry will draw many ideas from it, his interpretation of complexity will prevail.

In my experience, it's not difficult to propose a critical thinking problem to GPT-4o that it cannot solve properly and/or accurately, that even a grade school student can. We're not even close to AGI as long as we're still fiddling around with backprop-trained networks just to see how much profit we can squeeze from them. A proper learning algorithm that enables a machine to learn *from experience*, in real time, will invariably be vastly more efficient than offline backprop-training - and far more robust and versatile, without an infinite tail of edge cases that don't fit into its training data set.

Thank you for sharing. from a different timezone, it is great to have the recording, will follow up with interest

This is cool, I have this idea that (adaptive correction) is a bilateral symmetry to (resistance against equilibrium). I used this in vector math when coding a simple mechanism for thrust correction in relation to velocity direction, in relation to the direction of the desired position (bilateral mirror of thrust in relation to velocity proportional to the vector direction of the goal). When there is intermediate goals, shift in algorithm of control is proportional to proximity of obstacle (which is another bilateral symmetry). It's like a nesting of equal and opposite obstacle->correction vector comparisons based of goal of proximal importance.

There is a problem with data loss in that neural network example (if scale is the focus) for (category narrowing). In small neural networks that preform high degree information processing in order to narrow to one single outcome, a lot of information is lost (0,1 cat or dog 1, 0 outcomes) and this is fine on the narrow goal. However, if the information processing has to navigate a larger set of goals, and multiple networks are losing vast amounts of information that would otherwise be integral to the larger ensemble, the network must scale to be much larger to handle (lack of leverage [back propagation loss of nuance in favor of exaggerated certainty]) loss of data that would have otherwise been useful in a ensemble context where the goals are more interrelated, connected, and more numerous. Backpropagation becomes a less viable adaptive mechanism for larger systems, and becomes more expensive strategy when the parameters scale and goals becomes interrelated and numerous.

This is so cool!

So many gems. Altered my understanding of different fields. Phenomenal

Finding patterns in animal communication and understanding them. Will it help designing better robot?

Thanks

Amazing conversation!

A great conversation!

My cat, Smokey, recognizes herself in the mirror. Every night she sleeps in my lap, and there is a mirror in front of the bed at her eye level, so she frequently checks out her own reflection and will groom herself afterwards on her own belly, which the mirror allows her to see better. She can also see my face in the mirror, and she will frequently look into the mirror at the reflection of my eyes when I speak to her, so she not only recognizes her own reflection but mine as well!

Congratulations and thank you for continuing the dialogue about intelligence in its many forms. It is always a pleasure to hear from the Sante Fe Institute and your inspiring guests and scholars.

Thank you to the Santa Fe Institute for inviting us! This discussion was truly enlightening, and we're genuinely thrilled about the future of animal communication. It's amazing how much it can teach us about our place in the web of life.

Thank you🎉

Great discussion!!

We need more understanding of swarm intelligence, and what models result in emerging phenomena on network level that is not apparent on local level. In machine learning, I believe this would bring us to a new level of complexity/intelligence - social interactions bring forward a faster / "better in some ways" knowledge, our current models (I mostly hint at Reinforcement Learning, not LLMs, but LLM agents would be the same) do not have common sense which may come from no "common knowledge" if there is no one to share the wisdom with. I think if humans were not social creatures, we would not have much common sense either, that emerges from society and adapting to society. To my understanding, we do not have realistic, general frameworks for having compact agents whose behavior collectively emerge to something great. (Of course, we have particle swarm optimization and models for bird flocks, but none seem to really "emerge without a designed rule".) And because these agent models are still not compact enough to not be computationally expensive, especially when running a large network of agents (hard to paralelize, as agents interact), not much research goes into swarm and social intelligence in computer science...

❤-it

Excellent presentation. Homeostasis has long interested me and that functions on a similar (albeit more complex) way, simple mechanisms based on simple rules producing an emergent 'system' that addresses a whole series of problems.

You've never been vaccinated before with nRNA vaccine, Dr. West. BTW I enjoyed Scale a lot.

this is one of the most important talks ever given

rest in peace.

Thanks for posting this.

Great clarifying talk! Just the kind of things I was wondering about.

I don't know how the youtube algorith took me here, but I loved it. Simplicity and complexity

It’s too bad my tax dollars are being spent on this…

Awesome.

The psychology underlying Soviet Socialism and Corporate Socialism is actually very consistent (and not a little worrying), we do have the Soviet experience to learn from and that may help us predict the issues within the corporate world. AI has one great advantage and that is its bandwidth, given the baud rate of the humans by the time you have collected, verified, categorised and digested the information (requiring vast numbers of humans to interact) and then conveyed the summary (all that a single human can handle) to the selected 'decision maker' the matter in question is history. An AI system with access to data has enormous bandwidth and could process the data producing a result in near real time. Only an idiot would put an AI system 'in charge', its role is more akin to the civil service, gathering data analysing the trends and offering a set of potential strategies with both their outcomes and consequences. The humans could then pick which ever one made them feel best about their own nagging insecurities (which is what drives a great deal of human decision making). Issues would arise with the elements you cannot meter and processes which are essentially chaotic. As an example the rewards for compassion occur long after the event and are influenced by the countless interactions that take place in the meantime. There is considerable evidence to suggest that the benefits of compassion to all members of a group are marked and worth seeking but there is no way to directly measure compassion and no way to predict the outcome in such a chaotic system.

cat fail videos brought me here .

So what does self evidencing mean?

Most exciting talk I ever heard. Thanks Nick Lane for it and the books you have written. And thanks SFI for making this public and free.

"I recommend the videos featuring Dr. Hector Zenil posted on RUclips and his blog, for a more balanced view of the technical criticisms of Assembly Theory:

The question is not can you prevent science from doing harm (in the technical sense) Off course anything that does harm is conditional, and those conditions can be affected. That counts for making lethal compounds less lethal, or using alternative less violent farming methods, allowing for freedom of biodiversity within the human affected environment and beyond it. All these are obvious and in the public forum available. The real question is, can you prevent malicious intent or unintended consequences? Our biases, and the limits of our perceptions prevent us from making sure nothing bad will happen. Moreover scientists can affect public perception (or in the case of people like jordan peterson, misperception and misinformation with a phd to back it up) but in many ways they are powerless. The reason is, that the moment they move in the political sphere they have to compromise against their scientific responsibilities. That is : to not mix science and morality. To not move from what is, to what must be. So when scientists search methods to make sure citizens, politicians and the like follow their advice, they move out of the realm of science into politics, that is : affecting public opinion. Scientists have the responsibility to make clear this distinction. Science within it has not the imperative that one has to care about other beings, humans, animals, ecosystems... Nonetheless to have moral sensibilities (which often leaders lack, or loose in the game of cards and bluff and secretive violence) is essential to our humanity. To care for one another is part of who we are. Though it has s not a scientific imperative. It is no law of nature. Social science and moral sciences have their unique perspectives, but one cannot deduce from them how man must live, without sacrificing the art of being a scientist. In essence science has no relation to authority. It is fundamentally anarchist. And though great scientists existed in the past, we define not our reality based on their perceptions, but on how they help us questioning our current perceptions. Question the reality as it is, not as it was. In a sense all systems are systems of misinformation, as they imply that memory, that collecting patterns, repeated events is the path to knowing what is real. Yet what repetition is there but that of memory? Is the sun we see today the sun we see tomorrow? Is it even the same a moment from now? It is a matter of convenience, As without a little self deception, no knowledge can be gathered. There has to be some trust in repeatability. While being aware, that the deeper one goes into it, the less fundamental it appears, the more the contradictions between perception and experiment. So we dont really know what we say when we talk about protecting humans, climate, ecosystems. It is a matter of convenience we use such terms, and as human, as sentient being, i can say, it is valuable to do so. Because humans do not exist for sake of science, science exists for sake of humans, and one can never take the human element out of science. So i never idealize science. It can be used either way. Irrespective of safeguards. Those safeguards can be removed. People can choose to ignore the well meant advise, and even if scientists come with an army, they might meet a stronger army, led by scientist on the opposition. So first step is not do science, and then solve all problems. First is to nourish within people the sensibilities towards all living beings. Then whether they are educated in science or no t will not matter as much. It is kinda crazy, how sometimes scientists have impossible expectations from less educated people, as if they can replace 4 to 8 years of education with a blog post, a youtube video or a collection of data. Those people are not stupid. They just went a different path, often capable of things scientists can only dream of. And even if they have the education. It is pretty common amongst our greatest heroes, that they had strange beliefs, and convictions, which we would consider misinformation now. Brilliant scientists can make mistakes, yet we expect any dick or joe or amy to understand the intricacies of the scientific method, and the philosophy that questions and develops it even to this day, as we are neither all knowing, nor all capable even in the foundation of our sciences...

Mr. Interviewer, allow me to ask you the following: For what reason or reasons do you surrender to your dearth of self-confidence, manifesting, in veritas, invariably, inviolate toward ostentatious overwhelming onslaught of oratory overload in order to reassure your subject and/or your audience, but (with dauntless sincerity DO glance at your inner-mirror to realize) in reality it is, in fact yourself for whom you wish to recognize with silent cheers to fill the ears below the invisible laurel wreaths auto-bestowed by being big-mouthed, babbling and blathering blithely before bequeathing the electronic recording wand (known colloquially as “microphone”) to the seemingly sequestered subject to respond when, just before, you posed a frustratingly ugly, ceaselessly kicking, yard of useless taxiing over half-insulting masticated and near impenetrable spew you call a “question” - or twelve - to an emotionally indignant, obviously unamused Mr. Tymoczko, leaving him with little time, even less excitement to ANSWER your anti-pulchritudinous inquisitions ostensibly with the aim of eliciting wisdom’s grandchild, INNOVATION, mined from the aforementioned Mr. Tymoczko’s thoughts and theories of geometric musical mathematics.. of embracing new technology.. OF WHICH I AM INTRIGUED, THUS CLICKED ON THIS VIDEO TO HEAR YOUR GUEST SPEAK ON THE VIDEO’S TITLE: “Dmitri Tymoczko on The Shape of Music: Mathematical Order in Western Tonality”, but resulted in hearing instead YOUR audibly odoriferous emanations insufferably ululating for interminable temporal suspension in place of a simple rubric: let the expert in his own field speak to what he knows???

Great talk!

Very clear informative presentation. A great primer.

No one is actually ‘mad’ when you “say a screwdriver is life”; rather, you just undermine your credibility.

@47:00 "... in physics time does not exist... it's an emergent property..." That's such rubbish. What physicists are these folks? My guess is they're appealing to cosmological holography or gauge/gravity duality etc.? But none of the older or modern theories say "time is emergent". Some non-standard theories claim "space is emergent" (from entanglement structure) but *_none_* of them say time is emergent. Maybe Lee Smolin has infected their thinking with some sort of weird philosophy? Bottom line is: If you have lightcones and light rays, then you have a time dimension. What these weirdos might refer to is "no *_flow_* of time." But that's a totally different story. And mostly semantics: what do you _mean_ by a "flow" of time? You obviously do not know what you mean. Time does not flow. Perception of time seems like a flow, but that's perception, not fundamental physics. The perception is because we all learn to view and describe actual flows (of mass, energy, etc.) by clock measurements, which record time. The clock hands "move" or "flow" but time does not, time is just the coordinate. Coordinates never "flow" they just act as records.

Thank you.

She is amazing!

A conman who falsified their data. This pseudo scientist got 350K per year to conduct "research" in Canada. McMaster put him in paid leave while the investigation lasted 2 years until finally he resigned. Way to go McMaster, 400K in salary down the drain. White privilege at its finest. I bet McMaster would not do that to a BIPOC.

I'm so glad I found this video! I witnessed the synchronous firefly phenomenon while I was walking my little beagle, Link of blessed memory, late at night, maybe 6 or 7 years ago. A huge old oak tree around the corner from my house looked like it was covered in the flashing lights of a Christmas tree, which of course it wasn't (it was nowhere near Christmas and, anyway, the tree was far too enormous for anyone to decorate it in such a way). Link didn't seem to be all that impressed, but I stood transfixed. This was in Wilmington, Delaware. At last, I have an explanation. Thank you!

We Zombies can’t thank you enough. Brains!

A very good and informative video. A few issues though with over simplification; plant cells, just like animals cells have cognition, they solve problems in anatomical, and chemical space. This cognition is scalable and larger groups of cells have a higher computational and problem solving ability and are goal driven. This scales all the way up to the macroscopic level, with root systems, and plant morphology being agential and solving basic problems, such as reproduction via external agents, and seeking resources, etc. There is an awful lot of fact based peer reviewed data on this.

Yaaassss!

As an electronics engineer, I am wandering what synchronization does with objects getting synchronized regarding the amplitude? We see the phase is getting sync, but does it do something with the amplitude? From electronics amplifier design, a positive feedback is an unwanted effect, and most amplifiers have a negative feedback to get them very stable at fixed gain. If we want to make a oscillator we apply a positive feedback.

Brilliant