Looking for advice to build an AI Speech NPC world in Horizon
Hi everyone, I’m exploring the new NPC / AI Speech NPC features in Horizon Worlds. My idea is simple: Create a world with an AI-powered avatar/character. Enter through my Quest headset and be able to talk with it, getting real-time natural voice responses. I’d love to learn from anyone who has already experimented with this: What’s the workflow you used to set up AI NPCs? How did you connect speech recognition and TTS (external APIs vs built-in)? Any pitfalls I should know about (performance, scripting, UX)? In addition to general advice, I’m also looking for a developer who has experience building Horizon Worlds and experimenting with AI NPCs to help me create this. If you’ve done similar work, please reach out—I’d be happy to chat more about collaborating. Thanks in advance for any pointers!
[Blocking due to Game Editor crashing] Need fix
I wanted to flag two issues that are currently blocking development for many of us: Editor stability – The desktop editor often crashes, freezes, or kicks users out of a world without warning. This makes it difficult to iterate or debug. Avatar animation call – Calling playAvatarGripPoseAnimationByName consistently triggers a crash. Repro steps are straightforward: add the call in a script, press Play, and the editor shuts down. Both problems have persisted for about a week and effectively halt progress on projects. Any update or workaround you can share would be hugely appreciated—this functionality is critical for creators.Introduction and assistance please!
Hello everyone, first time poster so might be a bit of a long one. Jan 1st I began a self taught journey into Meta Horizons. I am Ex -British Military and Teacher wanting to create learning environments for my colleagues using VR technology, specifically in the field of Therapy. I have been building a small world and have made what I consider to be moderate gains given my lack of knowledge and skill but I am enjoying the process so that's the main thing I suppose. I have been relying quite heavily on AI using CoPilot, Grok, Chat and even the in built AI on the editor but I am finding increasingly that they aren't seeing what I am seeing quite often, so their suggestions (albeit said with huge confidence on their part!) are incorrect (unless its me just need understanding them and the editor as well as perhaps I think. One of my primary issues is that I am not in Desktop Editor but in fact Desktop studio. I have tried repeatedly to adhere to the AI instruction on correcting this but i just end up back in the same place which is with what I believe to be Desktop Editor and what they believe to be desktop studio. I think this disconnect is causing me to not be able to advance as quickly as I could. So my first question is: Can someone tell me if I am using the right software? I have downloaded and am using Worlds Desktop Editor (that's the title of the thumbnail with the "O" on the black background. AI is telling me I am in studio and not Editor and that's why their suggestions aren't marrying up with whatever software it is I am actually working with? Secondly, can I trust AI to continue to help me with this or do I need a different approach? Thirdly, I have an NPC created and I want to attach a text to speech audio file to give the NPC dialogue that can be interacted with i.e. on the click of the trigger the NPC starts speaking. Be great to hear peoples thoughts and I am more than happy to help out with the challenges I have encountered so far and overcome. Thanks in advance - Ant
