A New Kind of New Kind of Science by Owen Thomas Cunningham

Dear Owen,

To give you some background information on me. I received my doctorate in high-energy physics phenomenology. I was trained in theory, but actually wrote my thesis in applied theory: I used the ISASUSY/ ISAJET computer-based event simulation program to model the discovery and analysis of Supersymmetry at the (as yet not built) International Linear Collider (ILC). I later worked with Marshall Space Flight Center (NASA - Huntsville, Alabama) and used the GEANT event simulation program to model the behavior of their Scintillating Optical Fiber Calorimeter (SOFCAL) Cosmic Ray detector. My fellow professors amicably called me Dr. Cosmic Ray. I left the world of professional physics 10 years ago to join the business world. I now use computers to track financial benchmarks, sales, inventory, etc.

But my essay has nothing to do with computers - it's simply geometry, Lie algebras (you shouldn't poke fun at them if you haven't studied them closely enough - please read my paper, it introduces a simple 3-D geometry-based level of understanding certain Yang-Mills theories), and unification.

Your computer architectural ideas are interesting. My office assistant likes to talk about trits. She correctly argues that one way to improve computer speed is via machine symbol complexity. If we had a positive/ zero/ negative switch, we could define a base 3 trit, and could rewrite current base-2 bit machine words/ data into shorter base-3 trit machine words/ data. This idea might speed up computers slightly, but the investment into rewriting software might not be justified.

If I understand your graphs on page 2, your object architecture could speed up computers in at least three different ways: 1) you have moved from a single bus towards a large number of effective buses - this should speed up hardware, 2) the distance scales between CPU's, buses, and memory locations could be reduced - this should also speed up hardware, and 3) we could make more complex machine symbols, say base 256 (or as large as we care to go - based on how many orders and recruits we choose to use simultaneously) that could shorten machine words/ data in that new base - after considerable investment in software development, this should speed up software.

The process of recruitment seems inefficient - almost like Brownian motion, but could work in principle if the recruitment distances are short enough.

I don't understand the point of your philosophical interpretation and forward. Are we talking about emergent computers, emergent artificial intelligence, or emergent life? (Or are they all related at some level?) Regardless, you made several unnecessary and potentially inflammatory remarks in that section that do not respect other professional scientists. I will assume you did not intentionally try to provoke me/ us and remain civil about this discussion. By the way, computers cannot solve every problem. Please see Abhijnan Rej's essay on computing the ground state of string theory with 10500 parameters - even an emergent artificial intelligence could not accomplish such a task without some theoretical insight (or gauge theory chauvinism?) as to how to simplify the problem.

After reading your paper, I am worried that your program might be a worm that scores all gauge theory chauvinists with 1's.

Have Fun, but BE NICE!

Ray Munroe

--------------------------------------------------------------------------------

From: Ray Munroe Jr

To: Owen Cunningham

Sent: Thu, October 8, 2009 5:03:17 PM

Subject: RE: Hi from Ray Munroe

That sounds good, Owen.

Take Care!

Ray

--------------------------------------------------------------------------------

From: Owen Cunningham

Sent: Thursday, October 08, 2009 4:55 PM

To: Ray Munroe Jr

Subject: Re: Hi from Ray Munroe

Do you mind if I publish this email exchange as a post on my essay's forum thread? Now that you've articulated it, I'm wondering if a lot of people haven't made the same miscategorization: "I thought you were proposing an architecture that would be the next evolutionary step for modern computers." Now that we know our conversation isn't going to devolve into a slapfight (like the anonymous person who told me I should withdraw from the contest!), I don't mind it being available to all the other FQXi community members.

--------------------------------------------------------------------------------

From: Ray Munroe Jr

To: Owen Cunningham

Sent: Thu, October 8, 2009 2:23:49 PM

Subject: RE: Hi from Ray Munroe

Hi Owen,

OK - I won't launch the nuclear attack. I'll play nice, because I prefer it when everyone plays nice. I'm sorry I overreacted. I haven't scored your paper yet. I still like your ideas on page 2. Maybe I need to reread the end of your essay. It was thoughtful of you to offer your program.

I think I understand your feelings of being ignored because of being too closely associated with a category. I also feel that the "pure theorists" try to ignore me because I my thesis involved as much programming and simulation as theory. But someone has to bridge the gap between theory and experiment, and Monte Carlo-based simulation packages are the most accepted way to do that. Computers are relevant to physics, and I thought you were proposing an architecture that would be the next evolutionary step for modern computers.

My theory is a fairly complex multi-dimensional lattice, but I tried to break it down into understandable components. Many of the "pure theorists" (excluding my friend Lawrence Crowell) snub their noses at a geometrical approach to a TOE, but I tried to make the components of this theory simple enough that one shouldn't need a doctorate to understand it. I think you need a doctorate to be biased against it.

I think that understanding TOE will not require fancier "gauge theory chauvinist" equations (my TOE should include all matter fermions and all force bosons - not just gauge bosons) or more powerful "emergent artificial intelligence" computers (I know it isn't straightforward stuff, but you should also try to read Abhijnan Rej's essay) as much as it will require a fresh insight.

I am curious about decisional branching. Perhaps a computer program could figure out the TOE without the cumbersome framework of axioms. The axiom might be "It's like this because that's what works".

I keep trying to push forward with my ideas because I love this stuff. Even if I win $1,000, I still have more time invested in these ideas than that small prize.

Have Fun!

Ray

--------------------------------------------------------------------------------

From: Owen Cunningham

Sent: Thursday, October 08, 2009 12:46 PM

To: Ray Munroe Jr

Subject: Re: Hi from Ray Munroe

Hi Dr. Cosmic Ray,

Thanks for taking the time to read my paper and comment on it. I really do appreciate it. I hope you don't mind that I'm responding directly to you rather than posting another comment on the web forum.

At the time I cooked up the phrase "gauge theory chauvinist," I thought it would be met with a certain grudging amusement; the same way that, in my rebuttal to an anonymous critic earlier in this discussion thread, I referred to myself as a "a lowly private sector code monkey who flips zeroes to ones for a living." Obviously there is some truth to such a label, but I also don't realistically think that's all I am. I figured "gauge theory chauvinist" would elicit the same bemused recognition in professional theoretical physicists. Instead, it seems to have really pissed people off, which I regret and for which I'm very sorry.

I realize that I am the fish out of water here -- I am someone with no formal training in physics (beyond a pair of undergraduate introductory courses) who has submitted a paper to a physics contest. I don't have the fluency with the vocabulary of physics to express my ideas in a way that physicists will find intuitively approachable, so I was forced to express them using the vocabulary of computation. (I don't even want to call it "computer science" since I have only slightly more formal education in that area than in physics. My understanding of computation is more intuitive, borne of having begun programming computers at age six, and not having stopped doing it for another 26 years. Both a chemist and a sculptor can discourse knowledgeably about the properties of marble; my perspective on computation is more akin to the sculptor's than the chemist's.)

Given these deficiencies, I have a few comments I'd like to make in my paper's defense:

The physics community is the one that first decided to crash the computer scientists' party, not the other way around. It was physicists and not computer scientists who coined the term "digital physics," and that area still seems to be a fairly young and undeveloped subspecies of physics. I view my paper as an attempt to contribute only to that particular subspecies, since I agree I am simply not qualified to contribute meaningfully to any other area of physics. The downside to this is that theoretical physicists who don't have any interest in digital physics lack the same intuitive understanding of my computational vocabulary as I lack of their physical, Lie-algebra-based vocabulary. This is why my paper has elicited so many comments of the form, "Wow, your mildly interesting paper seems to have made some good points, but nothing really relevant to physics." I wish someone who had as much history as I do with computation, and also as much history as you do with physics, would read the paper and comment on it.

The biggest advantage that computation seems to have over mathematics in describing reality is its much more natural accommodation of conditionality and decisionmaking. Mathematics has never seemed to handle decisional branching gracefully (or at least not as gracefully as it handles other constructs), whereas it is an integral, natural part of computation. Many of the behaviors we observe in the universe -- not just behaviors that fit in the physics category, but also behaviors in biology, chemistry, economics, meteorology -- seem to have a conditional aspect to them (in addition to other aspects like magnitude, direction, orientation, motion, etc.) Mathematics can model these latter attributes quite nicely, but seems flummoxed by conditionality. So this has led me to wonder whether, if we do eventually find a theory of everything, it simply can't be adequately expressed in purely mathematical terms. What if the theory of everything is better expressed in computational terms -- that is, what if the theory of everything is better expressed as programming language source code than as a bunch of equations?

The very earliest spark of my paper was when I purchased Seth Lloyd's book "Programming the Universe," and thought to myself, "This is the first time I've purchased a book with the word 'programming' in the title that hasn't contained a single line of sample code." To co-opt the terminology of pure mathematics, the digital physics community seems to have contented itself with producing existence proofs, but not constructions. That community seems to agree that "Yup, the entire universe could indeed be software," but nobody seems to have taken the next logical step, to say "OK, what might that software look like? How might its source code be constructed?" My paper offers up a starting point for exploring such possible constructions.

In your comment, you mentioned that I made "several unnecessary and potentially inflammatory statements." Aside from the phrase "gauge theory chauvinist," which I agree is unnecessary and potentially inflammatory, what other statements are you referring to?

Thanks again for your time,

Owen Cunningham

Hello Owen,

Since I just read the comments about your paper, which you just left on Stephen Wolfram's essay forum page, I must confess that given my background I too should give your paper a read sooner rather than later. You might find my contest essay interesting as well, but I have some other papers which apparently relate strongly to the theoretical territory of your essay. I will report back here, once I have read it, and then attempt to give you a fair review.

All the Best,

Jonathan J. Dickau

Your comment about digital physics re my comment, both in Wolfram's entry, deserves clarification.. Wolfram shows discrete simple processes' outputs are compatible with otherwise woefully complex continuous models. Blog "Dot Physics" routinely shows an Excel spreadsheet trumps equations. So digital physics in silico. Process biases perception. Mathematics is not empirical, not a science. Contemporary physics is woefully biased in process (elegant models) exclusive of product (useful prediction). A bench chemist demands yield.

Galileo offended the One True Church not by saying Aristotle was crap, but by observably proving it. Physics offends me by saying "we've got it locked" and denying a contradiction within orthodox theory. If local left and right shoes vacuum free fall along non-identical trajectories, no prior observation is contradicted. That augers discovery. Theory that does not accurately predict is wrong whatever its basis, structure, process, peer vote, or writer.

EP tests since Simon Stevin (1586) and Galileo Galilei (1638) are composition-based with zero net output. Composition is EP-inert. Calculation of Petitjean's CHI (normalized parity divergence) for atoms in a lattice is discrete, atom by atom, within a growing enclosing sphere. Calculation of Avnir's measure is global not local. Crystallographic space groups are qualitative not quantitative. Add the odd-parity Chern-Simons term in all quantized gravitations. They all agree, yet physics refuses to fire its SOP gun with a novel bullet. Physics does not fear pulling the trigger, it fears a wrong shooter hitting the target. Ignorance is not a form of knowing things.

Thanks for the prior post heads up.

I am immune to little pellets.

Uncle Al,

I think I'm impervious to pellets.

Sincerely,

Nephew O

Greetings Owen,

I just finished reading your essay. You got or did a lot of things right. I think you were reaching somewhat, at the end, where much of what went before is fairly solid. You describe an interesting new architecture, which I would guess you imagine computationally resembles or describes the universe. It's an interesting hierarchal computational strategy, which is in some ways more realistic, but it would need to be shown it's a more accurate way to model reality.

Well; I believe that the Computational Universe paradigm has a lot to offer. I have greatly enjoyed some of the papers by physicist Paola Zizzi, and think you might like her work also. A Scientific American article by Seth Lloyd and Jack Ng credited her with coining the phrase "It from qubit." (more on that in a moment) I coined a clever phrase too, in imitation of Descartes "It computes - therefore it is!"

But I really loved Rudy Rucker's book Mind Tools, and then later found out he talked about computational universe (an infinite fractal CA, no less) in print a few years before Wheeler's "It from bit." (and Wolfram was one of his inspirations) I don't know when Fredkin first started talking about "Digital Physics" but Zuse's "Rechnender Raum" (Calculating Space) was long before that. So some of those ideas about a calculational universe have been kicking around a while.

However; "It from Qubit" is somewhat strange, compared to digital computing, and is a definitive step beyond "It from bit." Let me first assert that every unit of form can act as a probe/computer. It can receive, store, process, and transmit information. And this information does not have to be in the form of discrete bits, as a Qubit encodes a range of values, which can be decoded as a collection of bits. However; not all of the information possessed by, or which can be transmitted by a quantum entity, is localized or local.

That is; the wavelike aspect of a photon, sub-atomic particle, atom, or molecule is not precisely localized. Locality is instead an attribute of particle-like nature. So any sort of network topology that views each node as independent will fail to encode that aspect of reality. If the universe actually employs the methodology of "Computational Universe" theories, in its working method or process, it is likely carried out in a Quantum Mechanical framework. Which implies that some of the information is spread out over processors. To some extent your 'recruitment' idea encodes this.

But a natural computing process, at the microscale (i.e. - decoherence in nature), results in a web of entangled entities where each interaction is a measurement of sorts, that entangles the two particles which have interacted. This continues as interactions propagate across a network of quantum entities, so that A is entangled with B, B with C, C with D, and so on. This results in A being partially entangled with D, but more importantly makes the non-local information get spread out. So things get interesting. Daryll Leiter's paper touches on this decoherence as computing universe stuff, although you would never know it without some background.

More Later,

Jonathan

Hello again,

I wanted to share a few thoughts I didn't have time for last night, then I'll read any comments and respond. First; there is a paper by Fotini Markopoulou that might interest you "Planck scale models of the universe" can be found at arXiv: gr-qc/0210086. And she refers to a subject of my special interest, Causal Dynamical Triangulations, that bears a passing resemblance to your computational view - employing what is called a Random Walk or Markov Process. In CDT, a figure called the 4-simplex (or pentachoron) is their 'Lego' block from which the fabric of space is constructed. But their strategy of assembly and determination resembles a cellular automaton.

The CDT papers of most interest are "The Universe from Scratch" at arXiv:hep-th/0509010 and "Reconstructing the Universe" at arXiv:hep-th/0505154. There also was an article in Scientific American, on this subject. This should give you enough background to see the ways their methods resemble and differ from yours.

Please understand that while I think you have a great idea, I feel it's not quite complete as it stands. Therefore; by seeing how it links up with the work of others who have gone down a similar road, you may gain important insights into what works and why that will help you refine your concept into something far more powerful - hopefully with some predictive capacity. I think you are already well on the way to expanding some of the paradigmatic elements that hobble Science, but I think you need an expanded vocabulary or toolset, to accomplish your intended task.

All the Best,

Jonathan J. Dickau

Hi yet again,

I just noticed, when I expanded out one of the replies above, that you also found Rucker's "Mind Tools" inspirational. Way cool! I've lots of papers and books to recommend, if you want to dig deeper.

Regards,

Jonathan

A Postscript,

Causal Dynamical Triangulations is a computationally intensive theory, which uses a Monte Carlo simulation to reproduce the dynamical aspects of nature. Their unit-cell is the 4-simplex (4-d analog of tetrahedron). It works because it's causal, with the rule being that the timelike edges of the figures must match direction.

The cool part is that their simulation shows fabric of space to be 2-d at the Planck Scale (appealing to Loop Quantum Gravity folks), evolving to 4-d spacetime at larger scales, moving through fractal dimensions at intermediate levels of scale. You will find that Quantum Einstein Gravity (Lauscher & Reuter) makes similar predictions (worth checking out).

All the Best,

Jonathan

Greetings Once More,

I thank you, Owen, for your thoughtful response to my remarks. I have written a couple of papers which might be of value in your quest for links between the computational world and the behavior of systems in the physical.

The first appears in Quantum Biosystems Journal, and is entitled How Can Complexity Arise from Minimal Spaces and Systems?. The second was intended and submitted for a special issue of Entropy, but was later withdrawn in part for being inconclusive or contradictory (there were also publishing fees). That paper is now on viXra, and is entitled Does the Non-Locality of Quantum Phenomena Guarantee the Emergence of Entropy?.

I later reconciled some of the paradoxes I was left with at the end of the second paper, and will be presenting the outgrowth of that next month at FFP10 (abstract attached) - a common basis for thermodynamic entropy and quantum non-locality.

I'll attempt to address any questions you have.

All the Best,

Jonathan

Whoops,

Forgot to attach the abstract for Do Thermodynamic Entropy and Quantum Non-locality Have a Common Basis? I'll be presenting this at FFP10 in November.

Best Regards,

JonathanAttachment #1: 1_Dickau1.pdf

Dear Mr. Cunningham,

I regret your misunderstanding of the theories that I will enclose under the title digital physics, or NKS in the terms you used for your title certainly in reference to Wolfram's theories. Their main universe property is not that it is digital, it is obvious the world is not binary in its representation and operation, otherwise not even full-color pictures would exist. What they mean by digital is not binary but discrete. It's quite unfortunate that being an enthusiast of digital theories you don't make a clear-cut difference between and confuse one with the other in order to present no other but what the better developed digital theories already suggest. I recommend you first to read the sources you are referring too.

Best regards.

I forgot to say that the fact that a discrete alphabet can always be represented by binary notation is not only beautiful but convenient, that's why the 'digital theories' perform their experiments and provide explanations using mostly binary notation. Not because they suggest everything is binary.

Last but not least, IMHO I think that even though your earth-ground description might be interesting, you might lack of some basic theoretical background that would have let you know that there is no sharp clear cut between hardware and software, so it is such a distinction what is rather whimsical.

Hello again,

I've a bit of advice, which you are free to ignore. You are more capable of adapting than some of the people who appear 'stuck in a rut' to you. I would not waste time trying to convince RLO or Al about much, as you can keep what you like and set aside the rest. For you; Oldershaw's findings beg an explanation, and that possibility is fascinating. Robert feels that the 'beginningless' aspect of his cosmology is accepted, some questions become meaningless. However; only someone like yourself, who needs to understand why a particular order exists (in the first place), can solve the problem he has avoided or evaded.

Therefore; if there is a congruent answer or answers, they still remain to be found and are unlikely to be discovered by the man who built the model. In Florin M's essay Intro, he states that although we may not understand what it is, there is likely a rigorous mathematical explanation for it. Now; while this shows a bit of the Platonic ideal - in this case the belief that there is an archetypal expression for which real systems are an approximation. But then he is careful to state that in Physics axiomization, we can't go straight down the same road as Plato. And Rob is correct to state that doing so is only a "Glass Bead game."

But Florin is elegant in his description of how to remain always on the realist side of the border with the Platonic ideal. If you read the comments on Florin's page by RLO, it would appear that he does not appreciate how skillfully Florin avoids falling into the Platonist trap. The idea in all of this is that one should not let the chosen emphasis of others prevent you from investigating the parts of the puzzle you find fascinating. If you should find some gems in the work of Rob or Al, utilize them and extend their results and methods if you can. You need not convince them that your way to find the answers about 'why' are viable, just build the model and see what it does.

If somehow you are able to find a unifying principle they have overlooked, because they are empiricists rather than theoretical physicists, you will be able to show the meaning behind their work - which they were not capable to explicate.

All the Best,

Jonathan