Date: Saturday 2nd September
Time: 10:55am
A whirlwind tour through the recent developments in AI. What brought us here, what can these tools do now, and what potential do they have to be practically useful? This talk will look at the abilities, applications and pitfalls of using Large language models, followed by Image Generation models and how to go beyond “Make a picture of X”, using additions such as Control Nets, LoRas and Textual Inversions to enable more specific control over image content and composition. The LLM side of things would cover using ChatBots as a development tool and for game content generation and discuss the possibility of LLMs embedded within games for dynamic content generation. The Image Diffusion side will talk about producing images beyond the plain “Make a picture of X”, by covering the wide variety of enhancements that have occurred in the past year like ControlNets, Lora’s, and textual inversion.
About the speaker
Neil is a game developer from way back, and hasdeveloped games for DOS and on the Amiga, sold shareware on download.com and ridden the tides of the GameDev industry as a contractor or a solo Indie all the way though to making Mobile and Web games. Neil has a habit of entering game jams, producing quite a collection of oddities since my first Ludum Dare entry in 2005.
