TE
TechEcho
Home24h TopNewestBestAskShowJobs
GitHubTwitter
Home

TechEcho

A tech news platform built with Next.js, providing global tech news and discussions.

GitHubTwitter

Home

HomeNewestBestAskShowJobs

Resources

HackerNews APIOriginal HackerNewsNext.js

© 2025 TechEcho. All rights reserved.

Show HN: Strange Binary – a cellular automata game

13 pointsby mkrecnyalmost 4 years ago

5 comments

gus_massaalmost 4 years ago
I need like 1 minute to notice the instructions at the top:<p>&gt; <i>K - MORE CHAOS | J - LESS CHAOS | P - NEW COLORS | R - RESET POPULATION | D - REDUCE POPULATION | F - INCREASE POPULATION</i><p>This kills my back button. What are you storing in the URL?<p>I couldn&#x27;t guess the model and if I press J and K then I sometimes get very different behaviors. Is there some explanation?
评论 #27334744 未加载
qubexalmost 4 years ago
What is the objective? I can see the controls and I’ve made a few interesting patterns but what is the thing I’m trying to achieve? What’s the gameplay significance of the graph? What exactly is it even measuring?
评论 #27336464 未加载
sackerhewsalmost 4 years ago
I opened it in a new tab. By the time I looked at the tab it was in full epilepsy induction mode. Thankfully I don&#x27;t suffer from it.<p>Perhaps don&#x27;t start the game until it receives a focus.
评论 #27336467 未加载
tantaloralmost 4 years ago
What&#x27;s the objective?
评论 #27336539 未加载
mkrecnyalmost 4 years ago
Below is an email exchange I had with Stephen Wolfram about Strange Binary in September &#x27;20.<p>Hi Stephen -<p>You changed and enriched my worldview when I read NKS over ten years ago in college, and I&#x27;m excited for your latest project. I imagine you don&#x27;t have time to answer emails like this, but wanted to share a side-project, and ask three questions.<p>Strange Binary is a &#x27;game&#x27; that allows you to explore the computational universe in an aesthetically pleasing way. It&#x27;s not clear what the goal of the game is for the player - but it has the side-effect of building some small amount of computational intuition in the player.<p>Essentially you navigate through 2D toroidal cellular automata, by toggling the system rule up or down randomly in &#x27;dynamism&#x27; (a parameter that reflects how many system rules mandate a change).<p>As you navigate, the system state (ie &#x27;on&#x27; cells) is ported into the new system. This is aesthetically fun, but probably un-useful. It&#x27;s sort of like having objects that couldn&#x27;t have formed in universe A show up in universe B.<p>If you have a chance to read this, or look at Strange Binary, I&#x27;d love to know if you have any heuristics for thinking about how to make discovered systems useful beyond fun and aesthetics. From the point of view of building intuition, what other system parameters would you expose to players? Have you ever trained something like a neural net with many steps of system data - with a view to something like pockets of reducibility across similar systems?<p>Thank you if you were able to read!<p>Myles<p>Myles ---<p>Thanks for being in touch, and apologies for a slow response…<p>About your game. I think it’d be useful to be able to see the effect of a “simple seed” for each rule. It’d also be nice to see the “spacetime” structure, rather than just the 2D+time version movie. I’m not sure that your universe A -&gt; B thing is bad; it’s kind of reminiscent of successive layers in neural nets.<p>Plucking systems from the computational universe and having them be useful is, I think, a core challenge of future technology. In a sense, my lifelong effort on computational language is about making a bridge between human goals and what’s computationally implementable. But how best to go from “computation in the wild” [e.g. in your game] to things we humans find “useful&quot; is a big (and important) challenge.<p>A few perhaps relevant writings: <a href="https:&#x2F;&#x2F;writings.stephenwolfram.com&#x2F;2011&#x2F;06&#x2F;music-mathematica-and-the-computational-universe&#x2F;" rel="nofollow">https:&#x2F;&#x2F;writings.stephenwolfram.com&#x2F;2011&#x2F;06&#x2F;music-mathematic...</a> <a href="https:&#x2F;&#x2F;writings.stephenwolfram.com&#x2F;2017&#x2F;05&#x2F;a-new-kind-of-science-a-15-year-view&#x2F;" rel="nofollow">https:&#x2F;&#x2F;writings.stephenwolfram.com&#x2F;2017&#x2F;05&#x2F;a-new-kind-of-sc...</a> <a href="https:&#x2F;&#x2F;writings.stephenwolfram.com&#x2F;2019&#x2F;05&#x2F;what-weve-built-is-a-computational-language-and-thats-very-important&#x2F;" rel="nofollow">https:&#x2F;&#x2F;writings.stephenwolfram.com&#x2F;2019&#x2F;05&#x2F;what-weve-built-...</a> <a href="https:&#x2F;&#x2F;writings.stephenwolfram.com&#x2F;2020&#x2F;10&#x2F;our-mission-and-the-opportunity-of-artifacts-from-the-future&#x2F;" rel="nofollow">https:&#x2F;&#x2F;writings.stephenwolfram.com&#x2F;2020&#x2F;10&#x2F;our-mission-and-...</a><p>We’ve done quite a few experiments on finding pockets of reducibility using neural nets. The results haven’t been too impressive so far. Though as I think about it, I don’t believe we’ve really done cross-system studies of the kind you’re suggesting. In other words, are there “generically useful” abstractions&#x2F;reductions that work across lots of systems? Or, conceivably, if you have an “oracle” for one neural net, that can compute with one system, can “just the neural net” be useful for another system? But viewed this way, one’s approaching the question of whether inside a neural net one can have something like a CA. And usually the more sophisticated the computational behavior, the more difficult it is to train. You should try to figure this out!<p>Looks like you’re quite busy with tech projects, but I’d encourage you to consider our Summer School if you can spare the time. By the way, rather to my surprise, our physics project may provide a new, very practical approach to blockchain-like systems … so there may yet be a way to merge your tech and science interests….<p>Best wishes and happy holidays,<p>— Stephen