S6 E241 Machinima News Omnibus (Sept 2026)

In this news episode, Tracy Harwood and Damien Valentine explore three very different developments shaping machinima and virtual storytelling.

Tracy introduces DramaMancer, an experimental AI-powered storytelling platform developed through Midjourney’s Storytelling Lab. Rather than generating an entire story, DramaMancer combines an author’s characters, scenes and narrative framework with natural-language player input. Tracy considers whether this represents genuine player agency or merely a more sophisticated illusion of choice—and asks where authorship lies when stories are co-created by the writer, player and AI.

Damien responds by considering what happens when players push an AI-generated narrative so far from its intended path that the story can no longer progress, recalling the infamous research-order bug in X-COM: Terror from the Deep.

The discussion then turns to a recent panel featuring pioneers of Second Life machinima. Tracy reflects on how early creators transformed a social virtual world into a collectively maintained filmmaking environment through experimentation, shared knowledge and community initiatives such as Machinima Mondays. She contrasts this socially driven model of innovation with DramaMancer’s attempt to lower creative barriers through automation.

Finally, Damien discusses Reallusion’s new iClone RTX renderer, which promises ray-traced rendering inside iClone without exporting projects to NVIDIA Omniverse. He also explains his practical approach to facial performance capture with AccuFACE, including why subtle expressions, head movements and accurately performed dialogue can make such a dramatic difference to animated characters.

Key time stamps

01:00 – Damien and Tracy introduce the news episode 01:46 – What is DramaMancer? 03:12 – Storylets, player agency and authored narrative frameworks 05:50 – Who is the author of an AI-generated playthrough? 07:27 – The limitations and risks of AI-driven storytelling 09:06 – DramaMancer’s potential for machinima and virtual performance 10:20 – Damien responds: can players break an AI-generated story? 11:46 – The X-COM: Terror from the Deep research bug 13:37 – Why DramaMancer may be a more interesting use of generative AI 14:55 – Pioneers of Second Life machinima 16:26 – How Second Life became a virtual production environment 17:34 – Machinima Mondays and community-led knowledge sharing 18:33 – Second Life as a collectively maintained virtual studio 19:59 – Automation versus community as two routes into virtual storytelling 21:41 – Reallusion’s new iClone RTX renderer 23:27 – Missing particle effects and the challenges of NVIDIA Omniverse 25:55 – Expectations for the RTX renderer beta 26:19 – AccuFACE 2 and facial motion capture 27:13 – Combining full-body and facial performance capture 28:25 – Why subtle facial movements produce dramatic results 29:54 – Capturing the whole face, not simply lip-sync 30:34 – Final thoughts and episode closing

Credits -
Co-hosts: Damien Valentine, Tracy Harwood
Producer: Damien Valentine
Editor: Phil Rice
Music: Phil Rice & Suno AI