On the Nature of Time

I think it’s really interesting to see the similarities between what Wolfram is saying and the work of Julian Barbour on time being an emergent property. Both suggest a similar underlying ontology for the universe: a timeless, all-encompassing realm containing all possible states &#x2F; configurations of everything. But what’s really fascinating is that they reach this conclusion through different implementations of that same interface. Barbour talks about a static geometric landscape where time emerges objectively from the relational (I won’t say causal) structures between configurations, independent of any observer. On the other hand, Wolfram’s idea of the Ruliad is that there’s a timeless computational structure, but time emerges due to our computational limitations as observers navigating this space.<p>They’ve both converged on a timeless “foundation” for reality, but they’re completely opposite in how they explain the emergence of time: objective geometry, vs. subjective computational experience

I wrote up more or less the same idea ten years ago, but in what I think is a more accessible presentation:<p><a href="https:&#x2F;&#x2F;blog.rongarret.info&#x2F;2014&#x2F;10&#x2F;parallel-universes-and-arrow-of-time.html" rel="nofollow">https:&#x2F;&#x2F;blog.rongarret.info&#x2F;2014&#x2F;10&#x2F;parallel-universes-and-a...</a>

Do physicists think time actually exists? I wonder if someone has reasoned that time is an accounting method that humans have developed to make sense of their experienced change of systems.<p>Wolfram uses the words progression and computation a lot in his essay, but there’s an implicit bias there of assuming a process is deterministic, or has some state it’s driving towards. But none of these “progressions” mean anything, I think. It seems they are simply reactions subject to thermodynamics.<p>If no one observed these system changes, then the trends, patterns, and periodicity of these systems would just be a consequence of physics. It seems what we call “time” is more the accumulation of an effect rather than a separate aspect of physics.<p>For example, I wonder what happens in physics simulations if time is replaced by a measure of effect amplitude. I don’t know, tbh, I am not a physicist so maybe this is all naïve and nonsense.

Thought experiment on the nature of reality:<p>- In a much larger universe, write down in a log book every event to every particle at every instant, from the Big Bang to the restaurant.<p>- Put it on the fireplace mantle and leave it there.<p>This is basically a log of a simulation. It exists in much the same way as an ongoing simulation would, except that its time dimension isn&#x27;t shared with the simulating universe. But every observer within has had the same observations as if it did.

Is any of what he’s saying here, something he hasn’t essentially already said before?<p>The parts of this which were a little surprising to me (e.g. the bit about comparing time to heat, and the bit about running out of steps to do at an event horizon) iirc all linked to a thing posted a while ago?<p>I don’t share his enthusiasm for the phrase “computational irreducibility”. I would prefer to talk about e.g. no-speedup theorems.

I like thinking about hypergraphs that continually rewrite themselves. I&#x27;ve thought about it in terms of literary critique, or in &quot;compiling&quot; a novel. It reminds me of petri nets in a sense, where at any given moment, a character has a static model of the world, which can be diagrammed through a causal graph of conclusions and premises. Then, an event happens, which changes their understanding of the world; the hypergraph gets rewritten in response.<p>I&#x27;ve toyed with this with my own graph software when writing novels. It&#x27;s of course impossible to fully document every characters&#x27; model before and after every event that affects them, but even doing so at key moments can help. I&#x27;ve wished more than once that I could &quot;compile&quot; my novel so it could automatically tell me plot holes or a character&#x27;s faulty leap in logic (at least, one that would be out of character for them).<p>I&#x27;ve also tried the more common advice of using a spreadsheet where you have a column for each character, and rows indicating the passage of time. There you&#x27;re not drawing hypergraphs but in each cell you&#x27;re just writing raw text describing the state of the character at that time. It&#x27;s helpful, but it falls apart when you start dealing with flashbacks and the like.

Every time I read stuff like this I get super drawn to thinking about Sunyata* - In Mahayana buddhism, my understanding is that Sunyata doesn&#x27;t mean absolute nothingness or no existence, but all things are devoid of intrinsic, independent existence. Everything is empty of inherent nature because everything is interdependent... phenomena exist only in relation to causes and conditions. This relational existence assumes that things do not possess an unchanging essence... the ultimate sense, there is no fixed reality. What might seem like &quot;everything&quot; is actually permeated by &quot;nothingness&quot; or &quot;emptiness&quot; and that phenomena arise dependent on conditions, without intrinsic, permanent nature.<p><a href="https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;%C5%9A%C5%ABnyat%C4%81" rel="nofollow">https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;%C5%9A%C5%ABnyat%C4%81</a>

Seems like an appropriate post on a day when the Nobel of Physics was awarded not for Physics discoveries but for computer science...<p>But from Wheeler&#x27;s &quot;it from bit&quot; to Wolfram&#x27;s computational universes, the question is: where is the beef.<p>Now, there might be ultimately something worthwhile with the obsession with digi-physics. Mental models that seemed disparate may merge and become fruitful. It doesnt even have to be a fully formed toolkit. Newton&#x27;s invention of calculus was kinda sketchy. But he was <i>explaining</i> things with it, things that were not undestood before.

The thing that bothers me about the idea of the &quot;Ruliad&quot; is that it&#x27;s completely unfalsifiable. Even if we existed in a reality where true randomness existed, or computational irreducibility wasn&#x27;t a given, you could always argue that what we observe is just one finite local slice of that Ruliad where things <i>appear</i> to be deterministic (or computationally irreducible) due to our boundedness as observers.<p>It&#x27;s basically the modern equivalent of &quot;turtles all the way down&quot; because it pretends to explain the nature of reality by extending our definition of reality to fit within an all-encompassing mental model that only makes sense on a surface level.<p>Granted, the words &quot;universe&quot;, &quot;multiverse&quot;, etc. are insufficient in describing <i>everything</i> in a way that includes everything we currently want to include, but giving a new name to that abstract idea of &quot;everything&quot; isn&#x27;t itself a compelling argument to also say that everything exists as a static construct and that everything is computationally irreducibile and deterministic at a fundamental level. Yes, that makes sense in a physics simulation, but in reality, we don&#x27;t know what we don&#x27;t know. Placing the unknown in a conceptual box doesn&#x27;t imply that it&#x27;s now known.

Is there anything testable or falsifiable here? Otherwise it&#x27;s just preaching beliefs.

I really think that Wolfram&#x27;s descent into fringe science has hurt a lot of well meaning people that don&#x27;t know better and think that because he&#x27;s developed useful software that he should be listened to in these domains.

Time and space probably belong to consciousness, rather than the real world. The objective &quot;true&quot; reality may be utterly incomprehensible in its complexity, but we can imagine a &quot;slice&quot; of that reality that arbitrarily defines space and time so that the interior of that slice follows some reasonable rules. That slice of reality can be thought of as a high-level consciousness that defines rules of our physics. Other slices of the same reality are possible, GR-like or QM-like, including those that are computational and discrete in nature. One universe, but many interpretations. Within each slice of reality, it may be possible to define smaller subsets of reality, corresponding to smaller consciousness, down to the human or even more primitive levels. So what Wolfram is describing may be true, objectively, to the observers of a computational slice of the universe, just like the MWI may be simultaneously true to the observers of the MWI slice of reality.

&quot;(as I’ve argued <i>at length</i> elsewhere)&quot;<p>Everything he writes is &quot;at length&quot;. This looks like an interesting read with good ideas but it is so long and has no structure that I gave up reading. It may help to give an abstract in the beginning of the article.<p>The problem with the treatment of time in physics is that we can only measure time intervals not the philosophical Time (with capital T). But physicists gladly conflate the two.<p>Mach said: Absolute time [the philosophical Time] cannot be measured by comparison with another motion, it has therefore neither a practical nor a scientific value.<p>Which means that all of the &quot;t&quot; terms standing for time in astronomical equations are for time intervals and tell us nothing about the philosophical Time.

Disregard anything Stephen Wolfram says about anything other than his Mathematica software. He&#x27;s a pretentious, arrogant twat who thinks he&#x27;s unlocked the keys to the Universe and is trying to convince the rest of the world of his brilliance.

Wolfram has always been difficult for me to follow. I think it&#x27;s because he tends to drone on, I don&#x27;t know why. I don&#x27;t think even he knows why. My understanding of what I have managed to listen to or read is that being who we are, we don&#x27;t process information fast enough in order to see much of what is around us, even while it is happening before us. An example is to take a minute under consideration, you can think about how long a minute is. It&#x27;s tangible to us. It&#x27;s not very long. But if we think about how long a femtosecond is, it is not tangible at all. We can&#x27;t experience a femtosecond. We can experience a whole bunch of femto seconds, but not just one. This is just one example of what I perceived the meaning of his thinking to be. Is that wrong, or so far off? Not only can we not experience a femtosecond, we will never be able to experience a femtosecond because our brains are simply not fast enough and aren&#x27;t built to exist at such a scale. If that&#x27;s what it means, then does that mean that he is referring to our ability to exist in certain scales, and our tendency to know the scale in which we exist? And, to exist outside of that scale, requires different computational parameters? Additionally, is this an extension of dimensions, just in time, not space? Does he differentiate between the two?<p>I know that the perception of scale has more to do with, well, perception, whereas computational irreducibility (as I understand it to be, anyway) is more of a function of natural processes....or THE underlying function from which all other functions stemming from that, are built upon. ... Right? Between that and perception of the scale in which we have evolved to exist in, it seems like they are at least closely related...<p>Some of what has been discussed here in the comments has me doubting my understanding, is the reason I ask.<p>To extend my question, could computational irreducibility help to explain why the Universe tends to &quot;recycle&quot; so many parts of itself? Is that some sort of telltale sign that when we see these patterns (golden ratio, fractals, recurring structures in naturee), we are looking at a fundamental aspect of the universe in some form, or it&#x27;s computationally irreducible equivalent, or is this to be determined?

Where it&#x27;s nowadays standard practice in science to conceive of time as the dimension along which events are tagged, I would suggest the opposite: process, as a sequence of events, induces time. But also in the modern conception, time is derived from atomic events produced by a nuclear source. So, fundamentally the two conceptions are the same, but the process conception allows for greater freedom in what the underlying process may entail.

Is this a guest writer? It doesn’t have the Wolfram tone at all. It describes a universe that isn’t centered on Stephen Wolfram, for example.

Wolfram article on the nature of reality.<p>Cellular automaton on the first screen.

I think Stephen at least dares to ask the question.<p>Here is a little thought-experiment on the Nature of Time.<p>You take the three body problem and you pick an initial condition and generate the trajectory of the three body from 0 to T by integrating through time with some numerical scheme like Runge-Kutta.<p>Now you do it again, and again, generating each time a &quot;universe&quot; of three-body trajectories. Doing so allows you to build a dataset of physically realist three-body trajectories.<p>And now the kicker : You train a diffusion model on this (potentially infinite synthetic) dataset. Once trained, to build a &quot;universe&quot; (aka 3-body trajectories) you only need to sample from this diffusion model. There is no more need to integrate through time. Past, present and future inside the universe just fold themselves into place in order to make sure the universe follows the time-evolution constraint.<p>When working numerically, both these schemes can theoretically be as accurate as desired (error smaller than any chosen epsilon), although the diffusion model seems to potentially necessitate more memory in toy model, it&#x27;s not evident as the universe is stored in a compressed fashion which necessitate less memory when the universe is no longer a toy model.<p>The underlying question I perceive from Stephen works are is whether it&#x27;s more efficient computationally to explore all possible universes simultaneously in which case time is a mere constraint you have to solve, or to generate each universe independently stepping through internal time.<p>Although it may seems to be the same (our perception only having access to a slice of the multiverse), as in the end you get in both cases a physically consistent universe, the nature of the sampling process change the distribution of possible states. It also opens the possibility of shifting across various universes, not that we would be physically aware of (the previous universe and future universe), but we would benefit by experiencing a &quot;better&quot; universe. It&#x27;s the same vibe of ideas which states that our universe has been fine-tuned for life to be possible.

I&#x27;m a big fan of Wolfram&#x27;s physics project, however, he seems to be confusing thinking about physics (computation) with the continuous and ever-changing substance of the universe itself.<p>Time is a human idea to grapple with the fact that everything is both continuous and constantly changing. Time is simply picking out from that continuous change a sequence of changes or state(s) that occur during a measured standard sequence of change, such as the earth making a single rotation around its axis (day). It helps us manage and refer to and measure both the order of changes and the duration of changes or states using standards.

SW is the Derrida of computation. More words to add more confusion than explain anything.

Okay. Time is a computation. Patterned or otherwise predictable computations can be performed instantly and thus are not time. Only results that can’t be precomputed are part of our perceptions. That’s what I got out of it.

Physics does not explain flow of time at all. If one films a thrown ball, physics can tell from few frames its speed or where the ball is on the following or previous frames. But it tells nothing about why, when see the film, we perceive the ball moving. Articles like the above misses this.<p>In fact there is no even notion of direction of time in physics. All physical models are time-reversible. And even if we observe violation of, say, CPT, in nature, it still will not explain while we perceive time flowing in a particular direction.<p>This is very well discussed in the book “Time’s Arrow” by Huw Price.

Ok, so after the article on time as ought to be an emergent property[1], here we go with time from a computational point of view.<p>Can we at least receive a definition of computation that is not somehow depending of time being a given, explicitly or implicitly?<p>Am I alone finding this a bit taking aback? Like this is not physics or even general philosophy but plain old theological focus on the prime mover.<p>[1] <a href="https:&#x2F;&#x2F;www.quantamagazine.org&#x2F;the-unraveling-of-space-time-20240925&#x2F;" rel="nofollow">https:&#x2F;&#x2F;www.quantamagazine.org&#x2F;the-unraveling-of-space-time-...</a>

Discussed in Permutation City

Its certainly interesting, though the language its couched in wouldn&#x27;t be found in any philosophical discussion on time. This is all to say that it deals with concepts that have been discussed in philosophy for a long time, and these insights wouldn&#x27;t be considered &quot;new&quot; to someone from say mid-19th century Prussia. Certainly the &quot;progressive unfolding of the truth,&quot; in qualitatively different steps which Wolfram adopts here as his concept of time is no different from Hegel&#x27;s concept of time and the movement of history. I would recommend, for anyone interested in this sort of thing, to just read the &quot;Preface&quot; to his <i>Phenomenology of Spirit</i>.[0]<p>[0]<a href="https:&#x2F;&#x2F;files.libcom.org&#x2F;files&#x2F;Georg%20Wilhelm%20Friedrich%20Hegel%20-%20The%20Phenomenology%20of%20Spirit%20(Terry%20Pinkard%20Translation).pdf" rel="nofollow">https:&#x2F;&#x2F;files.libcom.org&#x2F;files&#x2F;Georg%20Wilhelm%20Friedrich%2...</a>

Fascinating, but I really wish this work was being published as a series of papers in peer-reviewed journals. Otherwise it’s hard to take the work seriously.

Im curious about how this relates to deterministic time and the lack of free will.<p>&gt;Our minds are “big”, in the sense that they span many individual branches of history. And they’re computationally bounded so they can’t perceive the details of all those branches, but only certain aggregated features. And in a first approximation what then emerges is in effect a single aggregated thread of history.<p>Does this allow free will?

I don&#x27;t understand how computational irreducibility matters for the perception of time. Surely, even a computationally reducible universe could be so insanely expensive to predict that it wouldn&#x27;t matter?<p>I also don&#x27;t understand why our inability to predict the future is related to our perception of time.<p>Overall, my impression is that this is an essay in philosophy (i.e, devoid of any content) rather than science.

Surely Wofram deserves the Nobel as much as Hopfield and Hinton? Not for this stuff of course (which I doubt many take seriously), but because he also provided us with an amazing computational tool without which physics would be very far behind where it is today?<p>[And at least I knew his name already unlike our current laureates whom I just had to look up!]

&gt; If we were not computationally bounded, we could “perceive the whole of the future in one gulp” and we wouldn’t need a notion of time at all.<p>Maybe, if we assume we aren&#x27;t axiomatically bound, despite knowing that we are (but that knowledge is rarely in context, so we can only know it <i>sometimes</i>...once again: time...weird).<p>&quot;Thought is Time.&quot;<p>- Jiddu Krishnamurti

He literally only cites himself in that article…<p><a href="https:&#x2F;&#x2F;media1.tenor.com&#x2F;m&#x2F;v6Awsd0YO7IAAAAd&#x2F;metal-gear-rising-metal-gear-rising-revengeance.gif" rel="nofollow">https:&#x2F;&#x2F;media1.tenor.com&#x2F;m&#x2F;v6Awsd0YO7IAAAAd&#x2F;metal-gear-risin...</a>

&gt; At the lowest level the state of the universe is represented by a hypergraph which captures what can be thought of as the “spatial relations” between discrete “atoms of space”. Time then corresponds to the progressive rewriting of this hypergraph.

I believe it&#x27;s simply a unit of measurement we use to understand the movement or rhythm on which the universe operates, so it could be termed the &quot;progress of computation&quot; if that makes more sense but it&#x27;s all in the same effort.

You have to figure time would carry on even if nothing else was happening . . .<p>. . . at the time ;)

How would a bag shaped universe experience time? <a href="https:&#x2F;&#x2F;youtu.be&#x2F;FYJ1dbyDcrI?si=9Ga7PCeac4EV4Y4_" rel="nofollow">https:&#x2F;&#x2F;youtu.be&#x2F;FYJ1dbyDcrI?si=9Ga7PCeac4EV4Y4_</a>

The idea that time is tied to computation makes me wonder if everything we see as &#x27;progress&#x27; is just the universe showing us the loading screen percentage of the game of life.

Space is distributed and time is a centralizing force. The serial action bottleneck forces the brain, for example, to unify and send one action at a time. This is also replicated in LLMs that are distributed internally, but generate one token at a time. So time is like the force of centralization while space supports the distributed side.<p>These two tendencies are reflected in the exploration&#x2F;exploitation tradeoff. The exploitation part is centralized in language and culture, while the exploration part is distributed across the components of a system. They work together to achieve intelligence, both are needed.

Everytime this work of Wolfram&#x27;s comes up, I think the same thing: what this is more than anything else, is a tacit argument that the universe we inhabit and are structures&#x2F;processes within, is computed in a strong sense. I.e., that we are living in a computational &quot;simulation,&quot; the substrate of which is not currently accessible.<p>That he doesn&#x27;t come out and lead with this, I find quite peculiar. I&#x27;ve asked him about this in person and not gotten a less cagey response. I assume that is because he does not want his theoretic hypotheticals to be binned under &quot;simulation theory&quot; and his overall world view so categorized.<p>But I don&#x27;t see another reason to pursue this line of conjecture the way he does. And as I suspect that that premise is actually true, it&#x27;s all good IMO.<p>Unrelated directly, but certainly adjacent, is that at the intersection of simulation-theories and AI, is the premise that a computed person (i.e, an AI) is uniquely situated to &quot;jail break&quot; our own reality, to exist in the framing one. (And you know, maybe it&#x27;s turtles all the way down a la Flatland, so...)<p>As Douglas Hofstadter and Daniel Dennett foregrounded, a simulated hurricane doesn&#x27;t get you wet, but a simulated poem is a poem in every frame. So too travel entities defined well by computation.<p>A good reason, if we needed one, perhaps, to get on with the business of elevating ourselves into a purely computational embodiment, I think. I&#x27;d like to pop up a level and take a look.

So you can&#x27;t go back in time for the same reason you can&#x27;t go left in Super Mario Bros.

Wolfram&#x27;s theories are still largely pseudoscientific, in that way they look a lot like string theory, minus the public funding the latter received.<p>Neither theory is really falsifiable : if new experiments are made that contradict the theory, it can just be adjusted to fit the new observations. As a consequence, those theories are unable to make any kind of prediction about our reality, which makes them pretty much useless. No wonder this &quot;research&quot; was never published in any physics journal.

Computationally unbounded observers see more of the future but what of free will?

Did he tackle Lorentz invariance?

when you die, people say that your time has ended. Does anyone know scientifically speaking what happens to time for a dead person

Without even visiting the page I can predict what this writing will be about with uncanny accuracy.<p>1. Big words at the start - pretending to hack at a problem so big that just swinging the axe is a major undertaking<p>2. The prose slowly drifts to make less and less sense; words have no practical meaning anymore.<p>3. Simplistic images galore. Various plots via cellular automata and &quot;pretty&quot; images show things that have nothing to do with the topic and are only distant metaphors at best. Yet these images are the proof that it all &quot;works.&quot;<p>4. A nothingburger by the end. Leaves you wondering, why did I read all this?<p>Every essay by Wolfram is the same.

Almost like time is the stack and space is the heap.<p>Meh. Almost.

I think he&#x27;s a quack trying to torture an explanation of the universe out of his pet theory that uses a lot of words to say simple things but doesn&#x27;t predict anything. If &quot;time is what progresses when one applies computational rules&quot; then how is the order in which the rules are applied defined in the first place?<p>Computational irreducibility is a neat idea but i&#x27;m not sure its novel or something that explains the entire universe. My basic intro course on differential equations taught us that the vast majority of them cannot be solved analytically, they have to be approximated. I don&#x27;t know if the irreducibility idea is anything fundamentally different than saying some problems are hard, whether its non analytical equations or NP hard problems.

I guess I’ll just wait for Sabine to say something about this.

<i>groans in metaphysicist</i>

The web became trashed over a decade ago.