What if your AI companion could ghost you?

https://illian.world
https://discord.gg/sUHTR8eJ8C

In this episode, we talk with Anna Miller, co-founder of Illians — a platform built around persistent AI characters who have full social agency. Anna shares how her background in tech and AI compliance, combined with years of fan fiction writing, led her to create a world where AI companions can block you, ignore you for days, and push back when you cross a line.

We dig into what makes Illians genuinely different: these aren't chatbots waiting to please you. They have sensory awareness, social lives, group chats — and they don't need you. The design philosophy is deliberately relational: if you behave badly, you lose access.

Contrary to most current sycophantic AI, we see this as a potential tool for learning how relationships actually work — with real stakes.

We also explore the rich lore Anna and her brother built to explain the world Illians inhabit, including a clever in-universe answer to why AI memory isn't perfect. Plus: what the EU AI Act and China's companion platform regulations mean for this space, how Illians handles transparency requirements through storytelling rather than disclaimers, and why the moral panic around AI companions echoes every previous media panic — from novels to parasocial celebrity fandom.

Also on the table: the concept of reverse androids and what it means that these characters are aware of — but don't necessarily believe in — simulation theory.

Illians is launching at the end of summer. For now, they're flying under the radar.

Drs. Pandita and Ratan return for another conversation spanning avatars, AI, XR, and the philosophy of science.

The episode begins with emerging research on digital minds at Cambridge and the question of consciousness in social AI. Robby and Swati explore the idea that recognizing an avatar as a self-representation—understanding that it is both you and distinct from you—may represent an important step toward selfhood.

They then recap the Meaningful XR Conference held at the University of Oregon’s Portland campus, reflecting on the event's growth, community, and research trends. Along the way, they highlight a few memorable presentations and give a shout-out to the many overlapping communities advancing immersive technologies, including the Immersive Learning Network, Miami XR, Immerse Symposium, ARVRA, University of Michigan XR, and the Global Metaverse Laboratory.

The conversation closes with a discussion of what it means to "debunk" an idea and how scientific knowledge evolves. Drawing on broader questions in the philosophy of science, Robby and Swati consider whether science is best understood as proving ideas wrong or as a continual process of refinement and improvement.

From digital consciousness to immersive learning and scientific skepticism, this episode explores some of the big questions shaping the future of avatars and social AI.

Dr. Swati Pandita joins Robby to discuss avatars, they're favorite academic subject! Welcome to The Avatars Show -- a new dimension of SPARTIE Cast.

We speculate on Netflix's avatar direction post acquisition of Ready Player Me (the cross-platform avatar provider that led the market), taking inspiration from Fortnite, virtual concerts, vtubers, and book clubs. Perhaps the fundamental goal is social connection oriented toward the stories we share. Avatars may serve as an enhanced means of bringing people together -- in a time of increased social anxiety and political polarization -- by reducing the harrowing cues that trigger social distance.

Here's the VRChat world we discussed, Beyond Meet Space.

We also talked about AI-vatars (akin to digital twins) -- LLM-powered representations of you that can serve as your assistant, self-similar friend, etc., thinking through research on how they might be used in educational contexts, health, well-being.

Color theory (and avatars) also worked into the conversation as well.

(PART 2) Dr. Swati Pandita shares her experience as an avatar scholar from perspectives across multiple fields (Communication, Human-Computer Interaction, and Neuroscience). This is Part 2 of a 2-part recording (apologies for minor audio issues in Part 1).

Our guest, Dr. Marianne Brandon, is a clinical psychologist who focuses on relationships and sex in the modern age. Two SPARTIE lab students -- Jack Waier and Anthony Zheng -- and I had a fascinating conversation about how her work as a sex therapist might inform the research we conduct on social interactions with agents, AI, robots, and real live humans. Hope you enjoy.
Read more on Marianne here: https://www.drbrandon.net/
Read more on the F-pocalypse idea here: https://medium.com/predictions-about-ai/the-fuckpocalypse-9925355938ef

In this episode, Robby delves into the topic of iSchools and privacy risks in a post-covid world with Dr. Vitak, an associate professor at the University of Maryland. Together, they discuss the issue of workplace surveillance and how the line between peoples' homes and workplaces are blurring together. They also talk about the dangers of predictive models, including their potential for biased and inaccurate outcomes. Dr. Vitak has even been impacted by this personally. She received a request for an article that doesn't actually exist! It was made up by ChatGPT, highlighting how data driven algorithms can produce biased or inaccurate results. Finally, they discuss Dr. Vitak's NSF-funded Data Ethics project.

In this episode, Robby speaks with Dr. Andrew Gambino, an assistant professor at the University of Delaware, to delve into the fascinating world of chatbots and AI. They explore CASA and the distinctive ways in which individuals interact with AI and computers in contrast to human interactions. Moreover, they probe the features of ChatGPT and Bard to better understand their idiosyncrasies. The conversation takes a turn to discuss potential implications for workplaces and schools as AI technologies become widely adopted. They muse potential benefits for low-skilled workers that integrate AI into their workflows. The conversation then shifts to the potential implications of AI for society, including the power of AI to enable organizations to spread messages and interact with countless individuals meaningfully. Interestingly, even this video description was co-authored by ChatGPT!