Stephen Wolfram: Cellular Automata, Computation, and Physics | Lex Fridman Podcast #89

That's one of my long time interests and hobbies, which I write about on HN and discuss with other people frequently. I'm supposed to be doing something else right now so I'll quickly drop a few disorganized quotes and links here. (Sorry I didn't have time to be more concise!) A few years ago I ran across Max Bittker's beautiful "Sandspiel", which is a delightful cellular automata toy that simulates sand and other rules: [link] A few days ago I saw him tweet some amazing stuff that resonated with me, which then led me to discover what he's been working with Lu Wilson (TodePond): Sandspiel Studio -- user definable rules using a block based visual programming language. [link] "working on goth fungus kidpix": [link] Lu Wilson (TodePond): [link] Sandspiel Studio: [link] Sandspiel introduction: [link] Sandspiel Studio in 60 seconds: [link] Rainbow Sand: [link] Huegene: [link] Flower: [link] TodePond's Spellular Automata: [link] We had a great discussion on the Sandspiel Studio Discord server, where I posted some interesting links: Check out the Long Now Foundation talk by Brian Eno and Will Wright about Generative Systems: 10:39 minute excerpt of Will talking about cellular automata and demonstrating with Mirak's Cellebration: [link] 6:48 minute excerpt of Will demonstrating Spore Metaverses: [link] 6:29 minute excerpt of Will demonstrating Spore Creature Demo: [link] Entire 1:38:50 hour Long Now Foundation talk: Playing with Time | Brian Eno and Will Wright [link] Craig Reynolds

I am a huge fan of Sandspiel, which Max described in this article from 2019, and recently I was delighted to discover that he and TodePond have been doing a huge amount of wonderful work since then. What happens when you combine Sandspiel with a Scratch-like blocks based visual programming language that lets you look inside and see how rules work, tweak and modify them, and even define your own rules for different types of particles? And then form a community around it for sharing Here is Max's and TodePond's brilliantly ambitious visually programmable sequel, Sandspiel Studio! [link] Here's my profile, where you can play with the version of Max's flower growing rule that he shows here: [link] ...to grow underground potatoes and fancy flowers: [link] I've written more about Sandspeil Studio and related topics of artificial life, cellular automata, and visual programming, and quoted some interesting discussion with Max and TodePond from their Discord server (they actually already knew about most of this stuff, but they love it as much as I do), in the "Ask HN: What weird technical scene are you fond/part of?" discussion, in reply to api's comment about Digital Artificial Life: [link] api 67 days ago | parent | context | favorite | on: Ask HN: What weird technical scene are you fond/pa... Digital Artificial Life -- as in evolving program ecosystems, artificial chemistries or cellular automata that can manifest life-like phenomena, etc. Haven't done much with it in a while but w

I recently watched a podcast with Stephen Wolfram (few months old) where he runs a photo of himself through Wolfram Alpha "ImageIdentify" and it classifies him as a plunger with 50% probability and 8% probability it's a human [1]. He then goes on saying how he's working on systems that would allow a self driving system decide what to do based on if the collision detected is a human or inanimate object, etc. This was just one particularly ironic example of what I'm seeing with ML - we saw this level of accuracy in classification years ago - where is the enormous progress ? There have been pockets where ML shows it's applicable with encouraging initial results but after that it seems to me that we suck at perfecting those system and the progress has been incremental at best. The only exception I can see is stuff like self-play where the system can generate insane amount of training data. [1] [this video]