A Broader Audience for Cellular Automata

For general discussion about Conway's Game of Life.
User avatar
dvgrn
Moderator
Posts: 10610
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

A Broader Audience for Cellular Automata

Post by dvgrn » February 21st, 2019, 2:25 pm

EDIT: Moved to a new thread, after splitting off from the Accelerating Spaceships thread.
simsim314 wrote:I was thinking about kickstarter or patreon to make it my primary job. People are doing different things like writing a novels, or making board games. Why not spaceships? I only wonder how much people are sharing your and mine (and dvgrn's) ideals?
I think about this briefly every now and then. At least, when I first consider it, it seems like it would be nice to at least be able to take a few months off between consulting jobs now and then, and devote full time to some subset of
  • proving omniperiodicity,
  • building a quadratic-growth replicator that Golly can run away with (unlike the 0E0P metacell),
  • the next edition of the Life Lexicon,
  • a proper modern update to Golly's pattern collection,
  • a crowdsourced glider synthesis record-keeping system associated with the LifeWiki and/or Catagolue,
  • hooks added to Golly so that Mathematica can call it directly,
  • an extensible rule naming system for Golly where a temporary rule table or tree is auto-generated based on the "rule = " part of an RLE header,
  • all the common Life search utilities made cross-platform compatible, so that they can be compiled right along with Golly and called directly by Lua front-end GUIs.
  • about a thousand other worthwhile things that clearly Somebody Ought To Do.
...Or Maybe Not
But then I think that anyone giving me money might actually have some right to expect specific results in a reasonable amount of time, which would mean I wouldn't get to work on whatever I wanted to! What if some new amazing discovery came up, and I was stuck in the interminable end stages of cleaning up a Life Lexicon release and couldn't pay proper attention?

So after due reflection, I usually decide that in the absence of a no-strings-attached grant -- heh, something along the lines of a MacArthur Fellowship would certainly be nice -- I'd personally rather keep things as they are, on a strictly volunteer basis. That way at least nobody can complain about whatever I manage to accomplish. I'd certainly throw in a monthly donation to a Kickstarter/Patreon account for someone else working on Life stuff full time, though!

Small Active Group, But Large Pool of Interested People
It does seem to me that there is some fairly large potential to spark wider interest in Conway's Life -- especially as the 50-year anniversary comes up between late 2019 and October 2020 (depending on the exact anniversary you're targeting).

There are probably some millions of people worldwide, on the fringes of what Eric Raymond would call the "hacker community", who inevitably discovered Conway's Life at some point. They were probably totally floored by the incredibly weird and amazing things that people had constructed as of 1990 or 2000 or 2010 ... and then they moved on, and today have no idea about the existence of Gemini spaceships, Snarks, Caterloopillars, 2-engine Corderships, Collatz 5N+1 simulators, Hydras, stable line crossers, minstrels, cameships, and so on.

Now, if you could get just a small percentage of these people to put in a dollar a month into a Conway's Life Patreon/Kickstarter, pretty soon you'd have a significant resource available to drive some interesting CA research. But that's a lot easier said than done, or else we could start dreaming up a non-profit organization with a volunteer board that dispenses grants of various sizes to people who complete research projects that are on the Official Wish List.

Better Advertising
Probably the simplest way to make recent Life developments visible to a larger audience -- and thus improve the size of the potential donation pool for our hypothetical grant-dispensing organization! -- is to keep up with news announcements on places like the conwaylife.com homepage and the LifeWiki homepage.

Maybe just as important would be to keep Golly's pattern collection up to date. Golly is still humming along pretty well at a steady 2,000 downloads per month and, recently that's been gradually increasing. It's dropped off quite a bit from previous highs, though: from 2012 through 2016 the average was about 4,000 downloads per month. Those downloaders are basically the target audience we're talking about.

Unfortunately the new interesting stuff, even when it makes it to the pattern collection, is usually buried in a sub-subfolder somewhere, with no really good advertising billboard space to tell people to "Look At This!" Changes.html doesn't quite count, since Golly doesn't do anything annoying like bring it up as a README after a new install.

The Biggest Collection of the Coolest Stuff
(Or It Should Be)
Maybe the biggest missing piece in the pattern collection has to do with the online Very Large Patterns page, which currently contains a Caterpillar but no Caterloopillar, and (strangely) a Gemini gun but no Gemini spaceship. It might be a good idea to start a new thread to try to collect donations of completed patterns for Very Large Patterns (with good comments included, since that's always the weak spot).

User avatar
77topaz
Posts: 1496
Joined: January 12th, 2018, 9:19 pm

Re: Accelerating spaceships

Post by 77topaz » February 21st, 2019, 4:43 pm

Ethicacha, an apgsearch-based game (see also its dedicated Discord server here), is also something else that has also been proposed to a) make CGoL more popular and b) bring in money for Catagolue and possibly other CGoL-related pursuits. It's been discussed quite a bit, but now we've gotten to a stage where our relative inexperience with coding is preventing us from making much progress with the actual code of the game. Volunteers would certainly be appreciated!

AforAmpere
Posts: 1334
Joined: July 1st, 2016, 3:58 pm

Re: Accelerating spaceships

Post by AforAmpere » February 21st, 2019, 5:22 pm

77topaz wrote: It's been discussed quite a bit, but now we've gotten to a stage where our relative inexperience with coding is preventing us from making much progress with the actual code of the game. Volunteers would certainly be appreciated!
I feel the issue is more the fact that we have no set ideas yet. Once that is done, then we would start programming it, but it doesn't make sense to at the current stage.
I manage the 5S project, which collects all known spaceship speeds in Isotropic Non-totalistic rules. I also wrote EPE, a tool for searching in the INT rulespace.

Things to work on:
- Find (7,1)c/8 and 9c/10 ships in non-B0 INT.
- EPE improvements.

User avatar
calcyman
Moderator
Posts: 2932
Joined: June 1st, 2009, 4:32 pm

Re: Accelerating spaceships

Post by calcyman » February 21st, 2019, 6:31 pm

AforAmpere wrote:I feel the issue is more the fact that we have no set ideas yet. Once that is done, then we would start programming it, but it doesn't make sense to at the current stage.
I agree with @AforAmpere -- you can't start constructing a building until an architect has proposed a design for it. (Or you can, but then find that lots of what you've built needs to be destroyed for one reason or another. It's exactly the same with programming.)

On the broader issue of trying to monetize CGoL or other cellular automata, which may or may not be worthwhile, it is necessary and sufficient to convince people that doing so is of value. Now, for people deeply interested in cellular automata, the intrinsic value of new discoveries is enough. (I believe that Dave Greene's stable reflector prize is proof of this.)

But to appeal to people beyond the existing CGoL community, there needs to be an extrinsic value: what benefit is there in funding CA research? I believe that there are benefits, some of them quite profound, and that we're only beginning to scratch the surface of the tip of the iceberg (to badly mix metaphors). I'm going to list a few here, but they're neither exhaustive nor disjoint (they overlap):
  • Better understanding of emergent complexity: even though cellular automata are too abstract and idealised to accurately model reality as a whole, there are certain phenomena (percolation, diffusion-limited aggregation, traffic flows, etc.) which are well modelled by cellular automata. The Catagolue project attempts to explore this in a generic way, looking at many random initial configurations in simple rules. (It also has a benefit to the CGoL community in general for finding new patterns and glider syntheses.) Stephen Wolfram's NKS has a similar philosophy, but with a more academic and less community-led approach (and a preference for 1D instead of 2D automata).
  • Artificial life: By this I mean attempts to model evolution in silico. Tierra and Avida are one approach to this, but they're not really cellular automata. Another example is the lattice artificial chemistries pioneered by Tim Hutton. (Don't tell him I said this -- he's excessively modest and will profusely deny the importance of his involvement in shaping this field.) I wrote hex13 in an attempt to allow people to explore these lattice artificial chemistries on a GPU, thereby simulating larger worlds for more timesteps and hopefully seeing new behaviour.
  • von Neumann universal construction: this is the big one. The creation of a working von Neumann universal constructor (which can fabricate its own silicon chips and solar panels from silicates harvested from a desert) would be a multi-trillion-dollar industry. One need only release a single example in the Sahara Desert and wait for them to reproduce exponentially; the result would be an expanding self-repairing network of computers which would render AWS/Azure/GCP completely obsolete. John von Neumann actually created cellular automata so that he could mathematically model machine self-replication, and cellular automatists have been optimising this ever since.
These fall under 'examples of why CA research can benefit science and technology'. In all cases, the benefit is quite distant.

A secondary tier, which has a more instant reward, is that of 'commercialisable projects which would contribute to CA research as a side-effect'; these include ethicacha and CA-based cryptocurrencies. Now that Catagolue has been around for 4 years (and 1 day!), most of the early bugs and 'oddities' have been resolved, and it's ready for the increased server load accompanied by such projects.

A third tier would be commercialisable projects based on CA software: cellular automata are a good backdrop for puzzle games, Panic Shooter being one example. I was also thinking that it might be possible to develop a VLSI design program on top of lifelib: Bill Gosper's 'compressed quadtree' idea underlying HashLife would also allow the efficient representation of immense circuits with billions of transistors into a relatively small amount of memory, and lifelib allows the efficient manipulation of such patterns.
What do you do with ill crystallographers? Take them to the mono-clinic!

User avatar
pcallahan
Posts: 845
Joined: April 26th, 2013, 1:04 pm

Re: Accelerating spaceships

Post by pcallahan » February 21st, 2019, 9:15 pm

calcyman wrote:On the broader issue of trying to monetize CGoL or other cellular automata
Is there a thread for this broader issue? I'm sure I'm not the only one who thinks about this occasionally. Right now, I have no ambition of getting rich, but it would be nice to have something to supplement retirement income one day. (Though in the interest of full disclosure, I haven't done significant Life work in about 20 years.)

I think it is remarkable that there is a community that, without any compensation, has engineered and optimized universal constructors in a specific, very simple CA. The significance is enhanced after recently rereading Poundstone's 1980s explanation of von Neumann's work and Conway's proof of universal constructors in Life. This is as real as work on VLSI CAD tools, the main difference being that nobody is paying real money for it. Could that change?

Anyway, it might be useful to start a thread on this topic if there isn't already one.

User avatar
simsim314
Posts: 1823
Joined: February 10th, 2014, 1:27 pm

Re: Accelerating spaceships

Post by simsim314 » February 22nd, 2019, 10:44 am

dvgrn wrote:But then I think that anyone giving me money might actually have some right to expect specific results in a reasonable amount of time, which would mean I wouldn't get to work on whatever I wanted to! What if some new amazing discovery came up, and I was stuck in the interminable end stages of cleaning up a Life Lexicon release and couldn't pay proper attention?
I was thinking about this question and my answer to it is that your relation to your donors is not necessarily based on expectations. I could be like a star in the big brother. I come from time to time capture my screen explain what I do, argue with people that I will do that because otherwise I'll do nothing and go work in some other place. Like this could be a show, and this kind of issues and reflections could be part of the show that some people watch to enjoy the reflective process as well. I don't have to be either slave nor free individual, the reality is actually somewhere in the middle between the two edges, and this is very complex balancing which happens in our reality now, in every day to day life. For me at least my arguments with the boss is not more pleasant but I have no control at all because the boss decides and it's not my company. So for me at least doing something based on my personal interest which I have for a while now, and proved myself to some extent, but doing this for money looks like a valid and reasonable opportunity in the 21st century artists market.

Now there is another topic which I think more important because based on it depends also the personal profit per each individual, is the community and the monetization.
pcallahan wrote:I think it is remarkable that there is a community
I wanted to stop here because in my opinion our biggest issue as a community is that we don't recognize ourselves as a group. We are currently individuals having a shared hobby, or shared interest. And although there is a community around this topic we're currently still feel much more individuals coming to the topic. Let me try to formulate it with example. How chess became from being a hobby to being a sport? A lot of people who are interested in chess started to gather physically together and focus around events. And we have news definitely, but we're missing the physicality (at least some sort of regular virtual meetings), this will cause self recognition as a group, and then we will need to try and draw attention to those meetings. Becoming physical and drawing attention is kinda the main point here, because our achievements deserve more attention. People know to appreciate simple findings like Sir Robin or the c/10 ship, people know to look at pi calculator, gemini, caterpillar 8 bit assembly computer and say this is amazing!

Which brings me to my next point
calcyman wrote:On the broader issue of trying to monetize CGoL or other cellular automata...for people deeply interested in cellular automata, the intrinsic value of new discoveries is enough.
This brings me to a question: today we can monetize art and different communal services much easier. We can try to create followers in youtube channel and ask them to contribute through patreon. Many artists are already doing this, but they didn't had to build the community from scratch. Yet this is a real thing nowadays unlike even 5 years ago. I'm not sure how much of the intrinsic value can be translated into dollars, but it's definitely a good point to start trying.

For example I was thinking about making a lecture course on CGOL. Just a standard curriculum course made by the top people from the field. I'm obviously might do it alone - but here is my suggestion, we all learned about CGOL in different ways, we all have stories, and we have discovered some important things in this field. Each one from his angle. We could make a standardized lecture course on youtube (or whatever). I'm even willing to organize it, or make it alone - but if we want to start working as a group we can make a communal lecture course about CGOL from A to Z, this should draw some attention. At least we will have something to base upon our physical community. I think we can all agree on the basic curriculum, and then maybe each one will tell more of his direction.
calcyman wrote:But to appeal to people beyond the existing CGoL community, there needs to be an extrinsic value:
- Very soon molecular computers which are basically hexagonal grids CAs will start to be real challenger to other sort of computations (quantum and classical electric). No one knows how to make computers from CAs. No one learned about universality and complexity in such designs. People think they have hexagonal grid in each cell there is N calculations we use k^2 grids this means we have N * k^2 calculation per second. They don't get that those calculation are extremely hards to make effective and useful - but we know at least the basis for this. We at least have the ability to know what we don't know. There is some theory existing in our community which is valuable, we know to design CA circuits and make from them useful calculations.

Another external useful thing is my third point
calcyman wrote:I was also thinking that it might be possible to develop a VLSI design program on top of lifelib:
pcallahan wrote:Poundstone's 1980s explanation of von Neumann's work and Conway's proof of universal constructors in Life. This is as real as work on VLSI CAD tools, the main difference being that nobody is paying real money for it.
I was thinking along those lines for a while now as well. But I think I've a design not only for chips to be able to simulate B3S23 with transistors, and on top of this simulation we will write some CA designs, but I think I can design some intermediate subset from both of the worlds using programmable units like in FPGA and convert them into some subset of circuitry in GOL, like another abstraction language. And we can think in this intermediate visual programming language - very similar to OO, that will automatically be able to compile to both, CGOL and transistors. Think about two dimensional assembly, each time you send gliders to the root, and from the root they go to the address and in each address there would some special class, and the approach to the address is infinitely large. I think I wrote the post about CA computer design using the root and tree for addressing and it can growth infinitely large and all the types would be infinite precision. I would gladly use lifelib, or integrate lifelib and golly in some manner, and use it. Also if someone will write a more general translation between both world, and my visual programming language will use only subset of this functionality this also be great enough. We should distinguish between tool and implementations. So my implementation of the chip is only one possibility, we can design 8bit X86 equivalent as well.
calcyman wrote:CA-based cryptocurrencies.
Not sure if you're familiar there is this ethereum project which is basically a city of contracts. There is also this solidity which basically some advanced version which makes it ridiculously simple to implement crypto currencies and run them in the etherium. I really know about this very little, but when I talked with people who know crypto well, they were impressed that we're having different stamps which have different values from our hash's and mentioned solidity.

Which brings to my previous point - we should be more in touch with other internet and crypto communities which made some name to themselves already. We should cooperate with them, and learn the techniques they implement to become popular. Obviously only making superemelly amazing things is not enough to make you popular (and to become monetized you need to draw attention).
calcyman wrote:von Neumann universal construction:
One of the things I was thinking about is something similar to the open source ecology project. We can start to take products one by one, at least products we need - and make it to reduced to basic parts and make sure we can produce every part. So we already can chose subset of factories and production lines existing today and make a minimalistic version of it. Now what's important is that this ecology would self sustained. So if I want to open a new society, I just start from this basic ecology which requires say 1-10K of people. And we can start to replicate. Each society can have it's own rules and paste of production and different types of economies etc. Completely decentralized society up to the minimal level of cooperation needed to have an ecology. We already live in self replicating ecology, we know someone knows to build each part of the reality around us, but we don't know those people, and we don't share their knowledge, not do we know of minimal self replicating ecology. In this model we would have an ecosystem that supplies all the cities with basic needs, and each city is capable to take care of itself in it's base capacity and then we can share our productions to each other thus always making sure each community is survivable on its own.

EDIT Another important point we can make about our community is a folklore. We can have different stories about different people. A little bit like what muzik tried to do, to tell his side of the story in his blog. But I think we can record those meetings and publish them online. A bit like Joe Rogan.

User avatar
pcallahan
Posts: 845
Joined: April 26th, 2013, 1:04 pm

Re: Accelerating spaceships

Post by pcallahan » February 22nd, 2019, 12:38 pm

I like where this thread is going, but it seems to have escaped both the subject heading and the forum.

TL;DR Could Life and CA interest become a big, known phenomenon like (a) Minecraft (b) Maker movement (c) FLL or Vex robotics (d) CoderDojo. What would it take to get there?
simsim314 wrote:Let me try to formulate it with example. How chess became from being a hobby to being a sport?
I think this is a great question, and one that is not asked enough. What would it take to get Life (and other CA hobbies) taken as seriously as chess? Personally, I see Life as a much more serious matter--and I see chess as a weirdly obsessive interest in a finite system with semi-arbitrary rules. (Note: I never advanced further in chess than learning how to play and that may color my views; clearly it is very captivating to those who study it.)

There are two possibilities. (1) Perhaps interest in Life has reached saturation. Those who love it know all about it, and know where to find others. (2) We could really promote ourselves more.

When I wrote my long mothballed webpage in 1995, I very much believed the latter. I still think that was true in 1995, when I was an unusually young member of a group, most of whom had been doing this since 1970! Now I see this website and lot of new names, so I am convinced that there has been a resurgence.

But it depends on what we're measuring ourselves against. It is certainly not on most peoples' radars, including kids who might develop an early interest, as chess players do, or musicians or athletes. Some already do, but the motivation is entirely intrinsic. You can't make a living at it, and even the recognition comes from a very small group.

It is possible for things to take off fast and unexpectedly, like Minecraft for example. It's less clear whether you can just decide to make something take off. It needs to be fun and accessible to beginners. If a program like Golly had been possible and available in the 1980s, for instance, it might have taken off like the interest then in Mandelbrot sets and other fractals, or possibly gone much further. Now it is sort of drowned out in the embarrassment of riches found for free on the web.

A big thing for me right now is the idea of realizing CAs in a physical, tangible form, e.g. as tiles or jigsaw pieces. I have posted about tiles that realize Still Life constraints as well as rule 110. It would be possible to design a set of 3D pieces that would fit together in layers to realize multiple Life generations (it's unclear if the set could be reduced sufficiently to be usable and manageable, but it might be possible to make it comparable to the variety of Lego pieces). I have often wondered, since people are attracted to the tangible, if this could help promote CAs. On the other hand, maybe it's just me. I am a little burned out on staring at a screen all the time and like the idea of things I can touch and manipulate (or eat for that matter--rule 110 cookie cutters would be easy to make).

For the above, we are back to the embarrassment of riches problem. Again, lots of free stuff in the form of 3D printer designs on the web. So much that it is hard for anyone to get any attention.

Or would the key be to turn it into a competition? One very popular activity is FLL robotics contests (and some competitors like Vex). These are sponsored by schools and parent volunteers with the thought that they are educational. Could the Life "community" (I thought there was one) sponsor a contest with specific aims, like the robotics contests, with tournaments and fun for people who may be very much at the novice level. I think that is a real possibility but it would be a matter of doing both the promotion and the grunt work.

Another popular non-competitive activity I have been introduced to as a parent is CoderDojo which provides an unstructured environment for kids to learn to program (though it is more like a class the way they do it around here). Could Life be promoted like that? Again, it would take strong interest, promotional savvy, and a lot of work.

User avatar
simsim314
Posts: 1823
Joined: February 10th, 2014, 1:27 pm

Re: Accelerating spaceships

Post by simsim314 » February 22nd, 2019, 3:12 pm

pcallahan wrote:Personally, I see Life as a much more serious matter--and I see chess as a weirdly obsessive interest in a finite system with semi-arbitrary rules.
I just remembered Dota2 and other esports are becoming more and more popular, and becoming as real as other sports.

I was thinking advertising the community as eScience. Just like eSports is getting crowd we should strive to get scientific credentials, and be recognized as scientists. We can start to work like regular professors if there would be enough financial support from patreon as an example. The things that professors do except of being payed, is making research, lead students and teach classes to find those students. Everything in the eCommunity is available to us, youtube, twitch, patreon, kickstarter etc. The only question in my opinion is whether I'll try to do it alone, or some top people would like to join and add value to the viewer (if the top of CGOL community will make a curriculum and teach the course together instead of only myself it could be a great start).

As for gamification for kids that you're suggesting my opinion is that we should talk and try to invent as many directions as possible, then we will try to pick the maximal expectation of profit from those suggestions. The big question remains in my opinion, how much it's my personal agenda and how much it's a common agenda for the community? I also respect our differences, and maybe we can create a "hive community" where each one of us is doing something else but we refer to each other, and use education and other materials from each other to capture more attention. I think the question is on one hand how much other people are willing to invest in this effort of popularization of CGOL, and the other question is if we have a general agreement something should be done into monetization/popularization direction, we can make e-meeting and talk it through, and this can be done only if people are willing to participate. I'm certain I'll do something about it - I hope I'm not alone in this one.
pcallahan wrote:You can't make a living at it, and even the recognition comes from a very small group.
This what people thought about playing video games and it was wrong but the reason is somewhat different. As gamers there are much more, yet the educational value from our community is way higher. It's definitely not on many people's radar, but also not black hole entropy - people still read the news, and know there were some research done in this field. At least those who are into popular science. There is a popularization effort which every field needs to be doing in order to become on the radar. I know many people who know about CGOL they played a bit with that, opened golly and saw the patterns and so on. People are into mathematics and computers, this is common. We just need to do something unusual for our community and think about something else other than mathematics in the pure sense. We need something like promoter.

User avatar
Moosey
Posts: 4306
Joined: January 27th, 2019, 5:54 pm
Location: here
Contact:

Re: Accelerating spaceships

Post by Moosey » February 22nd, 2019, 3:31 pm

simsim314 wrote:
pcallahan wrote:You can't make a living at it, and even the recognition comes from a very small group.
This what people thought about playing video games and it was wrong but the reason is somewhat different. As gamers there are much more, yet the educational value from our community is way higher. It's definitely not on many people's radar, but also not black hole entropy - people still read the news, and know there were some research done in this field. At least those who are into popular science. There is a popularization effort which every field needs to be doing in order to become on the radar. I know many people who know about CGOL they played a bit with that, opened golly and saw the patterns and so on. People are into mathematics and computers, this is common. We just need to do something unusual for our community and think about something else other than mathematics in the pure sense. We need something like promoter.
Indeed, my father, who was an animator (and now teaches animation), came up with an idea for an animated TV show about some people playing video games (basically). The response to his Idea was more or less “who wants to watch people play video games”? Now, the real question is “who doesn’t?” (Me— I’m not into pop culture.) If we could spread intrest in CA’s— just tell neighbors or whatnot— that would be great. This would be particularly useful with people around my age (I am twelve years old), because if they are interested enough in it, they’re bound to show it to their friends, of whom some amount will find CA’s interesting. If video games have taught us anything about society, it’s that once there is enough of an audience for something it will appear in prodigious amounts, so interest in CA’s will quickly spread once a critical mass of people interested in them is reached, because people will have a lot more CA-related stuff to stumble upon.
not active here but active on discord

User avatar
pcallahan
Posts: 845
Joined: April 26th, 2013, 1:04 pm

Re: Accelerating spaceships

Post by pcallahan » February 22nd, 2019, 4:20 pm

Moosey wrote:once there is enough of an audience for something it will appear in prodigious amounts, so interest in CA’s will quickly spread once a critical mass of people interested in them is reached, because people will have a lot more CA-related stuff to stumble upon.
Yes, and I had a feeling that Life had already appeared in prodigious amounts, which makes me wonder if history has moved on. I hope not. At one point, Martin Gardner's columns in SciAm were widely known and celebrated, and to a lesser degree so were A.K. Dewdney's Computer Recreations later on. It seems increasingly difficult to find published work of this nature, though there is a lot of it on the web, often of very high quality, and seemingly nobody is getting compensated or rewarded, even in terms of recognition. Case in point: I was trying to find out more about wallpaper groups and found complete plans for 3D printing models for them. Wow, I almost want to buy a 3D printer now. There are things like this all over the place.

So we're at this strange mix of a golden age of mathematical recreations in terms of what's available and seemingly a kind of collective "meh" from the larger public who might take an interest in them. Is it possible that only a few niche hobbyists ever want to work on Life compared to the number of people who would like to play video games competitively?

Who were all the people making Barnsley fractals on early PCs and is it possible there will not be nearly as many now because there is just a lot more readily available amusements on computers? With modern tools, designing Life patterns may not be quite the same as "building a cathedral out of dynamite explosions" (Poundstone's phrase) but there is a steep learning curve and those who just want to see some cool stuff happen will find a lot of other distractions.

That's my pessimistic outlook. However, I do think that we could do a lot better at promoting ourselves as a community, especially by leveraging the educational benefits. Having coached FLL robotics with my son (not very well I'm afraid) I can see how to set up a Life challenge as a structured set of specific tasks in the style of an FLL tournament. The hard part would be drumming up enough interest to get volunteers to participate in tournaments and schools to take it seriously as an enrichment activity.

To be clear, the "monetize" part does interest me a little, but I would really just like enough visibility that I don't have to mumble through a long explanation of my favorite hobby.

User avatar
Ian07
Moderator
Posts: 891
Joined: September 22nd, 2018, 8:48 am
Location: New Jersey, US

Re: Accelerating spaceships

Post by Ian07 » February 22nd, 2019, 5:29 pm

Could this conversation possibly be split off into a new thread? It's certainly interesting, but definitely a tangent from the original topic.

Personally, I'm a big fan of the "curriculum" idea mentioned above, as it would make CA seem less intimidating to a wider audience. This was actually mentioned on the Discord server at one point (note: this is a message link) in the form of a Numberphile-style YouTube channel but about cellular automata, which would certainly be an improvement over videos like this one which state the rules, show off a few things from the pattern collection while making no effort to explain what they're actually doing, slap on some generic music, and call it a day.

User avatar
pcallahan
Posts: 845
Joined: April 26th, 2013, 1:04 pm

Re: Accelerating spaceships

Post by pcallahan » February 22nd, 2019, 5:32 pm

(I completely agree about splitting off this thread. Is there an easy way to do it?)

To elaborate a bit on the FLL robotics comparison, I was just thinking about how to make a competition like the "robot game" part of an FLL tournament. This isn't very fleshed out but I was thinking of an analogy to the Field Setup

I can imagine a Life field with three kinds of areas. The first would be preset for the competition. It could vary in specific ways, such as a block in various possible positions and could include more active patterns like Buckaroos or stable reflectors. The second area would be reserved for player patterns. They could be anything that fits in the reserved windows and supplied, e.g. from a flash drive. The third (which I think would add to fun and interaction) would be the manual entry window. You would not be able to use any electronic storage but would be limited to anything you can draw in the window (which varies by human skill as well as size limitations). This would be the trigger for any ensuing events.

There could be a rule that the player-supplied patterns must be contained in the rectangle until triggered (in fact, that could be verified by software that analyzed the whole setup for stability just to make sure the players aren't doing something tricky with timers).

At "Go!" the player would enter the manual pattern. It could be as simple as a glider in the right place. Players could make their patterns robust enough to allow triggers at many positions or require the player to get it just right.

The competition would consist of specific challenges like "detect the presence of a block and output a glider if found." Obviously, it could be more advanced. You would have to work to build something that didn't just blow up, but it would be possible to draw from large libraries of patterns.

Has anyone made a game along these lines?

User avatar
simsim314
Posts: 1823
Joined: February 10th, 2014, 1:27 pm

Re: Accelerating spaceships

Post by simsim314 » February 22nd, 2019, 6:27 pm

@pcallahan I generally like the direction of your thought.

I think we need to invest more in gamification. I've some ideas about making game in golly to allow glider to be shoot from some "predefined" space and those gliders will become inputs from the user. It's possible to make a game API for CGOL. I'm not entirelly certain why people would like to play in CA unless we will implement some way to use universal constructors with other stuff. For this purpose it could be beneficial to switch from CGOL and use some more entertaining CA rules, where the replication takes 1k ticks. Anyway gamification is interesting direction. Making it more rigirous and more "sciency" is another direction. Making it crypto friendly is anotyer one. I think we have a lot of ideas the question is how to make something out of them? I think there is a profession for helping people and communities like ours. But for starters we need to have more clear view of the community itself. We know people who are activelly investing time to design patterns and to vote once a year. But who are those people which are reading all the topics? There is a lot of unknowns and I don't like unknowns.
Moosey wrote:once a critical mass of people interested in them is reached, because people will have a lot more CA-related stuff to stumble upon.
Agree and it's an interesting thought. But in order to do that we first need to become a real community. The place where the magic happens in my opinion is where people are becoming real. You see people watching video games final in theater in a hall. Why do they need to go into theater and watch people play video games? Couldn't they do it at home? The answer is they could but it's what makes the difference between a hobby and a sport. You need physically be present. Obviously for us making it into physical meeting is way beyond our current level. But we could make internet meetings, and talk about this stuff. We could be managing more as a community.

User avatar
77topaz
Posts: 1496
Joined: January 12th, 2018, 9:19 pm

Re: Accelerating spaceships

Post by 77topaz » February 22nd, 2019, 9:03 pm

simsim314 wrote:I think we need to invest more in gamification. I've some ideas about making game in golly to allow glider to be shoot from some "predefined" space and those gliders will become inputs from the user. It's possible to make a game API for CGOL. I'm not entirelly certain why people would like to play in CA unless we will implement some way to use universal constructors with other stuff. For this purpose it could be beneficial to switch from CGOL and use some more entertaining CA rules, where the replication takes 1k ticks. Anyway gamification is interesting direction. Making it more rigirous and more "sciency" is another direction. Making it crypto friendly is anotyer one. I think we have a lot of ideas the question is how to make something out of them? I think there is a profession for helping people and communities like ours. But for starters we need to have more clear view of the community itself. We know people who are activelly investing time to design patterns and to vote once a year. But who are those people which are reading all the topics? There is a lot of unknowns and I don't like unknowns.
I think some people have already thought of making Golly-like games - there's even a thread about a (working!) "CGoL shooter" on this site. More recently, we've discussed gamifying apgsearch with "ethicacha", which has the added bonus of immediately (rather than just indirectly) increasing the input to Catagolue and potentially being accessible to people not interested in CGoL - somewhat similar to CAcoin, but perhaps more accessible to a larger public.

User avatar
dvgrn
Moderator
Posts: 10610
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

A Broader Audience for Cellular Automata

Post by dvgrn » February 22nd, 2019, 9:56 pm

pcallahan wrote:(I completely agree about splitting off this thread. Is there an easy way to do it?)
There's a quick-mod tool called "split post", so I've given it a shot. If you don't like the new title, it can be revised. Here's the original topic.

User avatar
simsim314
Posts: 1823
Joined: February 10th, 2014, 1:27 pm

Re: A Broader Audience for Cellular Automata

Post by simsim314 » February 23rd, 2019, 6:07 am

Let me summon up the possible direction we are thinking in and let me try to explain what in my opinion is lacking still to become more profitible/popular/recognized.

We currently have three major directions.

1. The scientific/engeneering direction.
2. The gamification/entertainment direction.
3. Crypto currency.

1. The scientific approach

I think we all like CGOL because of unexpected complexity and variety of non trivial patterns it contains in very simplistic and natural CA rule. This obvious applied field to other areas. Our search paradigms, our expirience in self replication, our knowledge of how to analyze and create circuitry it's all mainstream science.

So what are we missing?

- Group of experts that meet physically or at least virtually in the same place and time and discuss things. We are doing it basically now and it's more pleasant this way that we can each in his free time to come and comment. Yet the percieved value of it, is low. Think of a chess game where each individual is making a move in his free time vs. serious game of 4 hours over the board. We are virtual and more advanced than most people in virtuality perception, but this is also making us being percieved wrongly by the larger public as weak community.

- Articles. Arxiv is now for a while an alternative way of doing science. We can as a community publish papers we find important achievments of the community in the last two decades. Emphesizing collaboration with other groups, and benefit to the scientific community. Publishing papers is an important part of scientific research. Obviously peer review will be done by us as by maybe other experts in computational theory.

- Curriculum, students and folklore. To create a group of experts it's not enough. In order to make it work we need to create curriculum, we need to make folklore and stories (I guess you heard the story about Galileo and Piza - this is not science, this is a folklore), we need to make it available to general publuc and make it meaningful to them. So that people will come and want to pay money to learn this thing. Part of the education is not only the material is the story we tell about the community. About the great people gathered here. And about the benefit of thinking in CA language analyzing other scientific topics.

- Collaboration with other groups and communities. A little by little the internet becomes full of different groups of interest. They start to change society. Like eSport, crypto currency, etherium which is decentralized city. We should meet with other groups, talk, discuss, find commonalities. Maybe we can help their interests or they can help ours.

- One of the most profound examples is electronic circuitry designers. We can collaborate in order to create chips based on common knowledge or maybe we can create some new computational paradigms, like no bits computer.

2. Gamification - entertainment - art

Unlike the scientific approach which tries to be serious and useful, this approach tries to emphesize the fun part existing in CGOL. The youtube with 2.4 million views shows that people like to see designs in natural CAs. They know to appreciate the beauty.

So what are missing?

- One big thing we are missing is buitiful graphics that makes splushes and realistic esthetic imagery. We need to accept this limitation and to find a way to tackle it. For example think of 3d buitiful creatures that in their brain run CGOL simulator which will in turn make people get interested in designing CA brains in order to play this game. The speed of the creatures will depend on the speed of their circuitry. So new discoveries in CGOL like snark will increase the speed and performance of these game creatures. One thing we can add is that the creatures can move very slowly as modern games are nother interesting waymany times based on long scale several action per day gaming. Unlike in the 90s gaming was full dedication, today people make few actions and wait.

- Another interesting approach I find a lot in pcallahan work. Having puzzles to solve and to play with. For example we can start from simple "complete the still life" puzzles. People can solve them but it's not that simple. It's like SAT solver. Another layer on top of this, would be physical puzzle to complete a SL like pcallahad calcyman and I developed.

- See plenty of other suggestion of gamification attempts.

- We should also distinguish between gamification and entertainment. Having something like Joe Rogan show or mathologer or numberphile for CA or CGOL would be entertaining but it will not use gamification strategy rather than some popularization done in entertaining available to wide audiance manner. This is viable option already mensioned as well. Just want to say people are thinking about gamification, but entertaining is another valid path.

- As some more abstract forms of currency and virtual property are emerging. It might be possible to sell our designs just like people sell CG art. I don't want to close the products, but we can make some transacaction based on caterloopillar just like people made transaction representing gold. Gold is obviosly rare but also designs like cateroopillars. They should have some market price. As long as there is agreement something is rare and hard to build and appreciated it should be accepted as currency of some sort. This is related to patreon. I think it's a nice way to covert art into money.

- On the cooperation side we might contact some of those guys already (mathologer or numberphile) and ask them to make show about CGOL.

3. Crypto currency

For starters I think this is a great approach. To fit in exactly the spot made by some other community. I hope calcyman will read about soliduty and etherium if he doesn't know about it all already, and will be able to state what is missing in this front.

Several more things about crypto.

- One thing that bothers me is that although beneficial to CA community, it's not directly representing the interest in CA. I mean people will mine stuff just to make money like it's with bitcoin today. Obviously it's better to mine CA than to mine something random. Etherium is trying to fix that. Yet still the hardworking miners will be those with bigger computers. Not sure this is healthy to the community but it's defenetelly worth trying.

- Another point is to convince the CA community that your CA coin = money. We are still give prices in dollars not in hash's of SLs. Something is problematic with our community still. We don't trust the CA coin, and I think for a good reason. I'm not sure I would like to have money to represent CA value by crypto design for currency. I would obviously prefer to be paid in SLs rather than local currency or even BC. Just saying there is some psychological barrier there. Like for example calcyman is owning more than 50% of the coin. I'm appriciative of his effort, but I'm not sure I would like to use monetory system where so much value owned by single man. We should at the least investigate this topic more. What makes algorithm a currency and maybe join the market of crypto currencies. I hope calcyman would be able to answer those questions.

User avatar
pcallahan
Posts: 845
Joined: April 26th, 2013, 1:04 pm

Re: Accelerating spaceships

Post by pcallahan » February 23rd, 2019, 12:02 pm

TL;DR we could also try something like Project Euler.
77topaz wrote: I think some people have already thought of making Golly-like games
That's true, but besides having a game, the idea would be to build a competition around it that engaged the attention of people who are completely new to Life.

Life challenges tend to be open-ended and only very interesting if they are "firsts" in some way, though people will sometimes compile large catalogs of similar things. It is also mostly non-competitive. Except for very high-profile questions, there is room enough for everyone to work on independent or complementary challenges. It's a lot like research.

For people who don't do this regularly, it may be more motivating to have a challenge that has a well-defined objective, is known to have a solution, and that thousands of other people are also working on. You don't have to give cash prizes. People will do things for badges and recognition. I think it could definitely work in an FLL robotics type of context, but you'd need many volunteers to turn it into a real thing.

Another idea, maybe requiring fewer volunteers, would be a Project Euler approach. Compile a set of challenges with solutions that can be tested automatically and then put up a web page in which people post their solutions. Like Project Euler, it could begin very simply ("Place these two gliders so both vanish after colliding.") The testing software would verify that the input consists of the required patterns and the result of generating them is as expected. It could move on to stabilizing oscillators, checking for presence of gliders and blocks, etc. The main difference between this and what Life hobbyists already do is that all challenges would have a known solution. There would also be badges and scores and everything people expect from a competitive environment.

Another model (though I am less familiar with it) is TopCoder.
Last edited by pcallahan on February 23rd, 2019, 2:03 pm, edited 1 time in total.

User avatar
pcallahan
Posts: 845
Joined: April 26th, 2013, 1:04 pm

Re: A Broader Audience for Cellular Automata

Post by pcallahan » February 23rd, 2019, 12:37 pm

simsim314 wrote: 1. The scientific/engeneering direction.
2. The gamification/entertainment direction.
3. Crypto currency.
These are all great ideas. (Though I question the social benefit of crypto currency, it can certainly help drum up interest.)

I don't know if this is a fourth category or can be rolled into 1 or 2, but I would add "education" somewhere. I don't mean advanced education as you mentioned in 1, but introducing the whole way of thinking to a K-12 audience with simple projects. Nearly 20 years ago, I was persuaded to write an article about Life for math.com (it's still up) and though it is really introductory, I still find people mentioning it to me. That suggests there is a wider audience.

Where I live (and many other places I assume) we see parents frantically trying to get their kids to learn "coding" (typically Scratch and later Python) whether or not the parents are themselves software developers or the kids are especially enthused about it. I personally don't think the entire world needs to learn how to program (and marvel that my own job has not been automated out of existence yet) but parents and schools take it seriously. This applies equally to robotics competitions. Schools show enough interest in this to sponsor their own teams.

The reason that parents can be convinced about programming is probably that they know people write software as a career. It is trickier to get the same parents to believe that their kids need to be able to grasp emergent systems and do a deep dive into one specific one (Life). Yet it makes considerably more sense to me than turning them all into Python programmers (as if the invention of the telescope meant that kids everywhere needed to learn to grind lenses). So what would it take to have local Life clubs? I mean that rhetorically, but you would need engaged parents and teachers and some knowledgable mentors (not expert-level or original researchers but aware of results and able to grasp the significance). What else?

Another reason schools and parents care is that colleges will notice it on an application. To get them to notice Life, we need the clubs. To get the clubs, someone needs to notice. It is self-perpetuating.

All of this is doable. It is not a pipe dream. That doesn't mean it will happen. I have time to contribute to an effort like that, but probably not the gumption to initiate it.
Last edited by pcallahan on February 23rd, 2019, 1:05 pm, edited 2 times in total.

User avatar
pcallahan
Posts: 845
Joined: April 26th, 2013, 1:04 pm

Re: A Broader Audience for Cellular Automata

Post by pcallahan » February 23rd, 2019, 12:59 pm

@simsim314 As for beautiful graphics and splashes, I think that helps (though we may have missed the window when screensavers were in vogue).

If we're going to add artistic effects to Life, we should not neglect sound.

I once played with adding sound to Life. I wanted it to be the "actual sound" (as I imagined it) so I considered patterns to act on a mesh of vibrating strings, plucked at each state transition, and used the mesh to solve a crude differential equation. You need to attenuate it because otherwise the waves will get more and more energetic. I was hoping to hear some interesting tones and maybe a doppler effect from an advancing and receding glider. My effort was disappointing, sounding a little like a lawnmower engine.

But you could make something a lot better, I think, by associating sound effects and music with known objects (one way to roughly detect an object is by the series of values in a single cell; this has the advantage of symmetry, though you get some false positives). An object originates a sound at its location and it propagates according to a speed of sound in the Life grid. You could place microphones at other locations and on moving objects. Or you could move the microphone according to the position of an avatar as if you're walking through this environment.

The above is unlikely to appeal to purists, since you're adding things on, but I think in the hands of an artist or musician with video game experience, the result would be worthwhile. It's unlikely to be a symphony, but it could at least be as pleasing as wind chimes or (in the context of a large design with oscillators) give the feeling of being inside a large, complex machine.

hkoenig
Posts: 258
Joined: June 20th, 2009, 11:40 am

Re: A Broader Audience for Cellular Automata

Post by hkoenig » February 23rd, 2019, 4:34 pm

If someone wants a project, try adding a UI (specific to the OS) that would display Catagolue results that were just reported that are "interesting". Or embed apgsearch in a UI so a user can see something happening that looks Life-like.

When I was doing the object censuses last century (I looked, and some of the files show creation date years starting with "19") I had a mode to display each generation, even though it slowed things down considerably. It was amusing to see the random soup stabilize and then bursts of activity as Gliders finally encountered something far from where they formed. (Added a timer to turn that mode off automatically, as I sorta used it as a screen saver.)

Long time ago I had an idea for a two-person Life game. Competitively use Glider collisions/constructions to try to achieve objects/patterns/clusters/constellations that are worth specific scores. When things die down/stabilize, the other player adds a Glider from infinity to start things up again. Never got far enough to figure out how to handle escaping objects like Gliders or Switch Engines.

User avatar
calcyman
Moderator
Posts: 2932
Joined: June 1st, 2009, 4:32 pm

Re: A Broader Audience for Cellular Automata

Post by calcyman » February 23rd, 2019, 4:53 pm

Another point is to convince the CA community that your CA coin = money. We are still give prices in dollars not in hash's of SLs. Something is problematic with our community still. We don't trust the CA coin,
What CA coin? There isn't one yet: that requires a blockchain, a network, and a framework for making transactions.
and I think for a good reason. I'm not sure I would like to have money to represent CA value by crypto design for currency. I would obviously prefer to be paid in SLs rather than local currency or even BC. Just saying there is some psychological barrier there. Like for example calcyman is owning more than 50% of the coin.
Maybe I haven't explained well enough: existing soups searched on Catagolue are not a cryptocurrency.
I hope calcyman will read about soliduty and etherium if he doesn't know about it all already, and will be able to state what is missing in this front.
I read the Ethereum paper when it was originally published. The main point is that it supports Turing-complete contracts, but that's not exactly what's needed for a GoL-based cryptocurrency. In particular, anything on the Ethereum network is bound by the following:
  • The proof-of-work, ethash, is no more useful than Bitcoin's SHA-256.
  • Both Ethereum and Bitcoin use elliptic curve cryptography, specifically the secp256k1 curve, for digital signatures.
If we want to make the coin actually use GoL as its proof-of-work, then it can't be an Ethereum token. It would instead need to be its own independent blockchain, level with Bitcoin / Ethereum instead of being implemented as smart contracts in either of them.

At this point, there's no reason to be unnecessarily restricted to elliptic curve cryptography. Since there's a very real possibility that a moderately large quantum computer could be built in the next 30 years, it seems slightly myopic to use a digital signature scheme that can be cracked using Shor's algorithm. I had a discussion with an expert on post-quantum cryptography, and he suggested using Dilithium for quantum-resistant digital signatures. This is the real reason why this CA coin would have an edge over other cryptocurrencies; it wouldn't be immediately broken by Shor's algorithm.

In the work-in-progress branch of the apgmera repository, I've included the Dilithium reference implementation and successfully verified each of:
  • Creating private-keys from passwords;
  • Digital signing messages with those private-keys;
  • Determining the validity and originator of a signed message;
In fact, the actual cryptographic part of the project is all in place: what remains to be implemented is a transaction scripting language and the networking infrastructure.
What do you do with ill crystallographers? Take them to the mono-clinic!

User avatar
simsim314
Posts: 1823
Joined: February 10th, 2014, 1:27 pm

Re: A Broader Audience for Cellular Automata

Post by simsim314 » February 23rd, 2019, 5:11 pm

pcallahan wrote:There would also be badges and scores and everything people expect from a competitive environment.
We can make it even more generic. Just state a problem a life grid with the problem, let people play with it and the answer could be in the a grid or in 1 out of 4 options - so we can ask abstarct trivia questions as well. Now we also ask people to rate the questions they liked. So we will open the question to anyone to ask, but we will also present people with more enjoyable question using their own report. We can also find a problem after which people stayed longer and let the problem have addictiveness rating. Now you obviously know about rating in chess - so for problems I'm not sure project euler is trucking the ratings but there similiar chess problem site where the problems have ratings as well as the solvers.

Edit I'm not sure if you grasp my previous point. I'll try to add some explanation and I'll address the sound later.

Have you heard about light bot? Now I have a way to make visual editing of the sort bot for CGOL. Basically I'm trying to develop such language that will be toyish but will allow to create geminis, caterloopillars and accelerating spaceships. But I'll take it to design code, and this code will be compiled into game of life state. But the toyish concepts will depend on CGOL. You will have a compiler but the real geeks will not use my compiler and will optimize everything to work in less ticks such that their robots will get more coins and prizes. And those robots are already running in very nice 3d graphics with sound and all the splashes. So the outside layer would be a nice mario style 3d game. But on the inside CGOL logic with some visual programming language that compiles into CGOL.

Now one of things I want to say possible to have is this grey goo, from which everything could appear in finite time. So this is kind of my currebcy. In order to make curcuitry in this place many times you will need make real searches in CGOL or othe CA my vosual language will support. This search takes time but it allows to support more and more complex layers and thus faster robotos that make you more points.

User avatar
77topaz
Posts: 1496
Joined: January 12th, 2018, 9:19 pm

Re: A Broader Audience for Cellular Automata

Post by 77topaz » February 24th, 2019, 12:42 am

pcallahan wrote:But you could make something a lot better, I think, by associating sound effects and music with known objects (one way to roughly detect an object is by the series of values in a single cell; this has the advantage of symmetry, though you get some false positives). An object originates a sound at its location and it propagates according to a speed of sound in the Life grid. You could place microphones at other locations and on moving objects. Or you could move the microphone according to the position of an avatar as if you're walking through this environment.
Wolfram Tones has a concept very similar to this, using 1D cellular automata to generate music. Also, here is a forum thread about someone who created music using CGoL.

User avatar
pcallahan
Posts: 845
Joined: April 26th, 2013, 1:04 pm

Re: A Broader Audience for Cellular Automata

Post by pcallahan » February 24th, 2019, 1:01 am

77topaz wrote: Wolfram Tones has a concept very similar to this, using 1D cellular automata to generate music. Also, here is a forum thread about someone who created music using CGoL.
Wolfram Tones is interesting. I'll have to play around with it a little more. But the key idea I want requires at least 2D. I want to incorporate speed of sound and distance attenuation. It's less about music generation than virtual reality. What I imagine is walking around a complex pattern with guns and glider streams as an avatar. As I walk closer to a working glider gun, I hear louder thumping (or something more musical). A glider or spaceship rushing towards me doppler-shifts as it passes me and recedes. A stable reflector is either silent or just a sub-audible chime until hit by a glider. As it activates, I hear explosive motion (or maybe something more like jingling bells). That sort of thing. I don't think it has been done. If it has I would like to hear it.

Peter Armstrong, oddly enough, has actually composed a piece based on a pattern I worked on many years back.

Sokwe
Moderator
Posts: 2644
Joined: July 9th, 2009, 2:44 pm

Re: A Broader Audience for Cellular Automata

Post by Sokwe » February 24th, 2019, 3:07 am

There's already a fair level of interest in mathematical distributed computing. Perhaps just increasing awareness of Catagolue would help to expand the community.

Also, this topic has been moved to the General Discussion board.
-Matthias Merzenich

Post Reply