Agenda
Game AI Through Players' Eyes: What Works, What Backfires
The same AI feature drew fierce backlash in one game and earned praise in another. Which in-game AI do players welcome, and which do they reject?
Drawing on 76,025 AI-related Steam reviews across 172 titles, this session maps how players perceive in-game AI along eight dimensions. Every finding is illustrated with real player reviews from shipped games and becomes a design decision that attendees can apply to their own AI features.
To make these findings immediately usable in production, the session introduces an 18-item diagnostic tool for screening AI features before launch. A worked example shows how the tool profiles the strengths and risks of an AI feature.
Takeaway
Attendees learn which characteristics of in-game AI are associated with positive and negative player reactions, along with the design implications of each. They also learn how to run a pre-launch check of the risks and strengths of an AI feature with an 18-item diagnostic tool.
Intended Audience
This session is intended for game designers and UX designers who shape in-game AI features, and for producers and product managers who decide whether and how to adopt them. No knowledge of AI or statistics is required.