Simulating Evolution with AI: Predators vs Prey

In this video, I build an AI predator vs prey simulation in pure HTML, CSS, and JavaScript, where evolving neural networks learn to survive, hunt, and adapt using NEAT (NeuroEvolution of Augmenting Topologies) — a real example of artificial intelligence, evolutionary learning, and emergent behaviour coded entirely from scratch.

Watch as cyan prey blobs wander aimlessly at first, slowly learning through generations of trial and error to dodge predators that get smarter with every mutation. Meanwhile, predators evolve hunting instincts — flanking, coordinating, or even behaving completely randomly — testing the limits of what AI can predict, and what it can’t.

This isn’t just a coding demo — it’s a visual experiment in survival, evolution, and chaos. You’ll see how patterns emerge, how randomness breaks predictability, and how simulated intelligence evolves when faced with impossible odds.

Each generation represents a new balance between order and unpredictability — as the prey learn to evade, and the predators evolve to outsmart them.

We’ll cover everything step by step:
• Setting up the AI predator–prey simulation in JavaScript
• Designing dynamic, glowing visuals and smooth motion physics
• Implementing the NEAT algorithm to evolve smarter behaviour
• Teaching agents to make survival decisions based on inputs
• Adding different predator archetypes — hunter, flanker, controller, random
• Exploring how unpredictability breaks learned AI strategies

By the end, you’ll understand how evolutionary algorithms can create intelligence from nothing — and why randomness is the ultimate test for any adaptive system.

If you enjoy projects that blend AI, programming, and creativity, subscribe to Programmetry for more videos like this — from evolving AI creatures to generative art, neural networks, and physics-based simulations.

🎥 Subscribe for more videos like this:    / @programmetry  

🤖 AI Experiments Playlist:    • AI Experiments  

Timestamps:
00:00 Introduction
00:18 Phase 1: The Basic Predator
03:16 Phase 2: Multiple Basic Predators
05:22 Phase 3: Advanced Predator Types
08:02 Phase 4: Predators Become Unbeatable
09:51 Conclusion

Song: The Dead
Composer: John Tasoulas
Website:   / john-tasoulas  
License: Free To Use YouTube license youtube-free
Music powered by BreakingCopyright: https://breakingcopyright.com

#ai #aiproject #aicoding #coding #programming #javascript #machinelearning #reinforcementlearning #neuralnetworks #evolutionarylearning #artificialintelligence #webdevelopment #codewithme #aiprogramming #creativecoding #simulation #aigame #browsergame #neuralnetwork #deeplearning #aiart #visualsimulation #programmetry #aicode #aiforbeginners #aiagents #ailearning #aialgorithm #aiprojects #codetutorial