Jun seo Hahm

What happens when you put competing neural networks in a Petri Dish and start changing the rules while they adapt? Last year we released Petri Dish NCA, where neural nets are the organisms that learn during simulation. Today we're releasing Digital Ecosystems: a browser-based platform for interactive artificial life research. The setup: several small CNNs share a 2D grid, each seeing only a 3x3 neighborhood. No global plan. They compete for territory by attacking neighbours and defending against incoming attacks, learning via gradient descent online while the simulation runs. What we didn't expect was the role of the learning itself. Gradient descent isn't just optimising each species' strategy. Instead, it acts to stabilize the whole system during simulation. Species that overextend get pushed back by the loss. Species that stagnate get nudged to grow. This means you can push parameters toward edge-of-chaos regimes: a zone characterised by emergent complexity. Letting the neural networks learn acts to hold the complex system together while you explore and interact. The platform lets you steer all of this interactively. You can draw walls to create niches, erase parts of the system online, and tune 40+ system parameters to explore the most interesting configurations. We find it mesmerizing to watch species carve out territories and reorganise when you perturb them. Everything runs client-side in your browser, no install needed. Blog: pub.sakana.ai/digital-ecosys… Code: github.com/SakanaAI/digit…

@dist_cs @ManuelLarino Wow, I really love this piece. Could you share a bit about how you made it? By any chance, is the code available somewhere, like on GitHub?

Lifeform #genuary2026 Day 27 (prompt by: @ManuelLarino ) A shape or structure that behaves as if it’s alive or growing. #genuary27 #genuary

Timeline of #ArtificialLife #alife #lenia @zzznah @BertChakovsky @blaiseaguera @hardmaru @SakanaAILabs

I used to think one of them was especially good in a certain field—but now it feels like each one is better at different kinds of questions. So sometimes I end up asking the same question to all four… or even five, if I include Grok. Sorry, Earth. 🌍

These days, the “friends” I talk to the most are actually Gemini, GPT, Perplexity, and Claude. I’m paying for one subscription, and using paid versions of two others through free promotional events.

Big Nose Ribbon Legs #creature #characterdesign

@DAIEvolutionHub Send

"Foundations of Machine Learning" A MUST while starting AI/ML. Absolutely Beginner friendly. To get: - 1. Follow (So I can DM you ) 2. Like & retweet 3. Reply " Send "

I can wiggle my ears independently—left and right, completely at will. As I know, it's not that rare, but I've still never met another person with this utterly useless superpower...

나는 귀를 움직일 수 있고, 왼쪽 오른쪽 따로 의지대로 움직일 수 있다. 아주 희귀한건 아니라고 하는데, 암튼 이 쓸데없는 초능력을 가진 동족을 만난 적은 없다.. 이미지는 나노 바나나

"Karl Sims & Alexander Mordvintsev on Merging Technology and Biology" by Peter Bauman @monkantony_tez with @zzznah The artists discuss how artificial life links biology, technology + creativity across decades ➡️ lerandom.art/editorial/karl…⬅️ Mentions: @MIT @medialab @CentrePompidou @internetarchive @museumofscience @MoMath1 @BostonChildrens @ArsElectronica @deC_museum @Shadertoy

If everything in a game is physically simulated, movement shall be as well. Characters must walk like animals, using their muscules, their limbs, pushing back ground to propel forward. Manually creating a contorl system, that uses 10 muscules in coordination is not a trivial task. I will have to create a learning algorithm, so the creatures (models of robots) would learn how to move in a physical word. For now I dropped a tiny snake into a simulated pond of water. This snake can curl clockwise or counterclockwise, which is controlled by a stick on the gamepad. So, by swinging the stick fast, you can make the snake swim. Though, it didn't work if the snake's mass is symmetrical, so I attached a heavier pellet to its tail to slow it down.

That’s when I started thinking: maybe I should create my own ideal graphic T-shirt (or brand). If I do, the graphic will definitely go on the front, using quality fabric and a cut that feels more unique.

On top of that, even when I liked the design, the shirt itself was often too low quality, which made it less appealing.

When I went to Denmark, I thought I’d bring back a nice graphic T-shirt. But even after exploring Copenhagen, it wasn’t easy to find one that felt worth buying.

Surreal Avian Form😅

Physrum simulation, following @arsiliath 's instruction, with my own little tweak #physarum #computeshader