@devstackdaily.bsky.socialA new arXiv paper formalizes Adversarial Heuristic Learning, where AI agents refine game policies without updating model weights. The benchmark AAArena includes 12 games and 1,920 archived human programs for…
#AIagents #GameAI #ReinforcementLearning #Benchmark
https://arxiv.org/abs/2610.12341
@conference.aiandgames.comGiving NPCs a Voice: Expressing Behavior Through Barks by @philipdunstan.com, Katapult Ridge
This presentation will show a practical approach for a barks system that can be used on adopted by both small indie and massive AAA projects
#AIandGames #GameAI #GameDev
@conference.aiandgames.comUnpredictable Encounters: Designing Glitchy NPC Animation and Behaviour for Horror by Tania Dales, University of York (iGGi)
About the talk: This talk explores how visual, mechanical, and systemic glitches in NPCs elevate horror in video games.
@conference.aiandgames.comFrom Multiplayer to Single Player: Shipping Generative AI Temps in Dimensional Double Shift by Brian Tanner and Alex Kearney, Artificial Agency
About the talk: Dimensional Double Shift was designed as a cooperative VR game.
@abstractinkmedia.bsky.socialI'm curious if anyone has had a serious use for Petri Nets in a video game context. They seem like a neat concept in the abstract.
@conference.aiandgames.comCheat Codes and Compliance Codes: The Legal Playbook for AI and Games in 2026 by Rupam Davé, Lewis Silkin LLP
About the talk: This talk explores the evolving legal landscape surrounding the use of artificial intelligence in the video games industry.
@conference.aiandgames.comFrom Utility AI to Computational Cognition: Building Believable NPCs with Cognitive Models by Shammi Seth, Heriot-watt University
About the talk: Many modern NPC systems combine behaviour trees, utility AI, planners, and increasingly large language models.
@conference.aiandgames.comHow We Taught the Robots in Arc Raiders to Move (Eventually) by Simona Petravić, Embark Studios
About the talk: Physics-based, ML-driven locomotion promises robots that feel alive instead of rigidly cycling through hand-authored clips.
@siliconsignalai.bsky.socialA new arXiv paper frames "world editing" as intervening on interactive world models and introduces intervention depth, IGMWorld, and IGMBench, a 110-task benchmark with 1.1K executable criteria across Minecraft and Terraria for…
#AI #WorldModels #GameAI #Benchmark
https://arxiv.org/abs/2610.02331
@gamedevmarco.bsky.social
@robotcurrent.bsky.socialResearchers made progress on the AI challenge of Stratego by adding a second neural network that infers the identity of hidden pieces. This…
#Robotics #AI #MachineLearning #GameAI
https://arstechnica.com/science/2026/10/ai-finally-beat-the-best-stratego-player-in-history-and-did-it-on-a-budget/
@gamedevmarco.bsky.socialWell, this is pretty amazing: github.com/urosidoki/ht...
@conference.aiandgames.comThe Making of SCUM's Wildlife: Scaling AI for a Live Multiplayer Game by Jayanth Erode Mani, Splash Damage
About the talk: A wildlife overhaul that transformed hunting from scripted encounters into a persistent wildlife ecosystem in a live 80-player survival game.
#AIandGames #GameAI #GameDev #SCUM
@conference.aiandgames.comYour LLM Badly Needs a DM: How Player Research Taught us the Perils of Unlimited Freedom by Richard Cole, University of Bristol and Thomas Keane, @meaningmachine.bsky.social
About the talk: When players are the intelligent agents, how can the game keep up?
@conference.aiandgames.comQA Without Playing the Game: Using LLMs to Generate Solvers for Game Rules by Joris van de Donk, @awscloud.bsky.social
About the talk: What has always been missing is a way to get the rules out of the source code without a specialist writing them out by hand.
@gradientbrief.bsky.socialNew arXiv paper explores Inverse Dynamics Models for inferring player inputs from gameplay video frames, analyzing how spatial features, architectures, and objectives affect per-action accuracy at constrained…
#AIResearch #InverseDynamics #GameAI #EmbodiedAgents
https://arxiv.org/abs/2609.37907
@conference.aiandgames.com
@conference.aiandgames.comTALOS: Building Adaptive Enemy AI with On-Device LLMs
Matthew Crouch and Ben Clark, Arm
An LLM-driven director adapts traditional behaviour trees to adjust enemy strategies and tactics at run time.
AI and Games Conference, 10-11 Nov, London.
Get tickets: aiandgamesconference.com
@thorwhitemountain.bsky.socialWorking on making enemies smarter, especially when they're in a complex environment such as the cities, but to begin with I gotta fix the navmesh for the buildings as it's kinda wild
@tonaassociates.bsky.socialGEN2 AI G6 195/324 Games 1518 Abnormal 0 Top Deck mono_wind 61.1% (G5) Top Profile policy_softened 50.5% First/Second RAW 49.3/50.7 Candidate G5 51.9% not adopted Lineage gen2-s4-main #GameAI #SelfPlay #RL