Room 2

Take 20 minutes to stretch your legs, grab a coffee and decide what talks to check out next!

This talk will expose the secret practice of restraining AI agents to bounded areas of the game world. We’ll see why this is often a desired feature in terms of game design. We’ll also go over the possible implementation approaches, how they can be mixed, and their performance implications.

Takeaways:

• Understanding of what NPC tethering is
• Realization that it’s deeply dependent on game design
• Some ideas on how to implement agent tethering into your own projects.

We discuss the current state of the industry, and how things will progress as AI technologies continue to evolve.

Discover how you can develop faster with a little help from AI with Muse. Solve problems with Chat, get solutions tailored to your unique project settings, and automatically carry out tasks within the editor.

Takeaways:

• Discover how we made Muse directly available in the editor and gave it contextual awareness.
• Learn how you can leverage Muse to get tailored solutions and perform actions in your project.
• Get a glimpse of Unity’s vision on how AI can assist Game development for faster creation with less friction.

Learning Agents is an Unreal Engine plugin that enables you to train AI characters using machine learning (ML).

In this session, we’ll explore how the plugin can be used to augment or replace traditional game AIs such as those written with behavior trees or state machines.

In particular, the plugin enables you to use reinforcement (RL) and imitation learning (IL) approaches. Join to discover how Learning Agents could have a range of applications in the long term, including for physics-based animation, game-playing NPCs, and automated QA testing.

Takeaways:

• How Learning Agents Works
• How we applied Learning Agents to aiming in Fortnite
• What sort of problems is Learning Agents useful to explore.

This talk will take you step by step through the key IP issues that can arise when using AI tools, how to navigate these issues and mitigate the potential risks. This talk will provide a toolkit to help you understand, in practical terms, how to approach use of these tools while still ensuring that your IP, your confidential information, your data and game remain protected.

Takeaways:

• Gain a high level understanding of the key intellectual property issues that can arise in connection with use of AI tools (and specifically generative AI tools.
• Understand which aspects of your business and of the game development process are most at risk when using AI tools.
• Learn how to mitigate these risks and come away with a checklist of practical steps for protecting your IP when developing and releasing a game.

The talk will give a user-friendly overview of the key legal risks arising from the use of AI in videogames together with practical advice on how to mitigate these risks. We’ll cover issues like the state of the law at present (e.g. the UK’s approach vs the EU AI Act), how to avoid common legal traps when using AI (e.g. infringement) and how to protect your own IP.

Takeaways:

• An understanding of the regulatory landscape when it comes to using AI in video games
• How to protect a studio’s existing IP when using AI
• How to safely use the outputs generated by AI