Video games AI

This is a portfolio repository collecting scripts, articles and implementations about AI in Video games.

Check my Reinforcement Learning and NLP models in HF

https://huggingface.co/jjmcarrascosa

Medium Articles written by me

Historical evolution of AI techniques

• AI in video games: a historical evolution, from Search Trees to LLMs. Chapter 2: 1950–1980
• AI in video games: a historical evolution, from Search Trees to LLMs. Chapter 2: 1980–2000
• AI in video games: a historical evolution, from Search Trees to LLMs. Chapter 3: 2000–2023

AI content generation for video games

• Unity Muse

Own algorithmic implementations

Heuristic-guided Dijkstra implementation in C++

Own implementation created during an interview process with Paradox

Own libraries

mindcraft: an LLM-eempowered-by-RAG engine for NPC

Use Zephyr 7B with a RAG-system containing lore and memories and past interactions of characters to model, using personalities+moods+motivations, tailored conversations with NPCs.

Projects in Unity

LLMs

Mistral 7B for CPU real-time content generation

Integration of Mistral 7B LLM into Unity, using LlamaSharp, to generate content (lore) for your games on-the-fly.

Behaviour Trees and Blackboards

Behaviour Trees with Blackboard singleton from scratch in C#, no plugins used.

Own games

A Lord of the Rings 2D tile card game in Unity, content generated with Dallee3 (images) and Mistral (text).

It's an ambitious game in progress with lots of complex logic, including Behaviour Trees and lots of complex logic.

Projects in Unreal

Navmeshes and Pathfinding

Simple Shooter

Simple Following AI: Toon Tanks