Innovating in XR & AI with Varun Siddaraju

Varun Siddaraju — XR & AI Researcher, Founder of VeeRuby

Exploring the future of AI, AR/VR, and human-centered innovation—building bridges between technology, research, and real-world impact. varunsiddaraju.substack.com

Episodes

  1. Oct 13

    The State of XR — From Hype to Human-Centered

    Note: This audio uses an AI-generated voice model of Varun Siddaraju’s voice, based on his original scripts. It is disclosed explicitly because this work is about trust, explainability, and human-centered systems. Season 1 finale. Recaps the throughline — context-aware spatial systems, the four problems (perception, spatial memory, adaptive behavior, explainability), world models as infrastructure, and “a demo is not a tool.” An honest read on the state of XR: hardware is arriving, the human-centered systems layer is the real frontier and barely started. Boundary held: active research, not products. Links:https://varunsiddaraju.com/ ·https://varuninnovates.com/ · https://veeruby.com/ #XR #AI #SpatialComputing #MixedReality #HumanComputerInteraction This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit varunsiddaraju.substack.com

    The State of XR — From Hype to Human-Centered

  2. Sep 29

    A Demo Is Not a Tool — Surviving the Messy Day

    Note: This audio uses an AI-generated voice model of Varun Siddaraju’s voice, based on his original scripts. It is disclosed explicitly because this work is about trust, explainability, and human-centered systems. “A demo is not a tool” is the honest lens for spatial computing: demos run on one user, one session, one happy path; tools survive the messy day — interruption, task-switching, recovery. Why polish and robustness can be inversely related, and how the four problems (perception, memory, adaptation, explainability) are really the integration test for surviving real work. Links: https://varunsiddaraju.com/· https://varuninnovates.com/ · https://veeruby.com/ #XR #AI #SpatialComputing #MixedReality #HumanComputerInteraction This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit varunsiddaraju.substack.com

    A Demo Is Not a Tool — Surviving the Messy Day

  3. Sep 15

    World Models Are Becoming Infrastructure

    World models — interactive environments systems can perceive, predict, and act inside — are becoming infrastructure, not features or products. Why the infrastructure framing changes the questions (ownership, consistency, inspectability), how it maps onto the four problems, and why the durable value is in the state layer. Boundary: “becoming,” not done. Links: https://varunsiddaraju.com/ · https://varuninnovates.com/ · https://veeruby.com/ #XR #AI #SpatialComputing #MixedReality #HumanComputerInteraction This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit varunsiddaraju.substack.com

    World Models Are Becoming Infrastructure

  4. Sep 1

    Explainability — If It Can’t Explain Itself, You Can’t Trust It

    Note: This audio uses an AI-generated voice model of Varun Siddaraju’s voice, based on his original scripts. It is disclosed explicitly because this work is about trust, explainability, and human-centered systems. Explainability is the fourth and keystone problem: the property that makes a fallible adaptive system trustworthy. Adaptation must be visible, reversible, and tunable. Why explainability fights the demo, why it’s a design constraint rather than a late feature, and why it holds perception/memory/adaptation accountable. Boundary: still an open design problem. Links: https://varunsiddaraju.com/ · https://varuninnovates.com/ · https://veeruby.com/ #XR #AI #SpatialComputing #MixedReality #HumanComputerInteraction This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit varunsiddaraju.substack.com

    Explainability — If It Can’t Explain Itself, You Can’t Trust It

  5. Aug 18

    Adaptive Behavior — When Should AI Step In?

    Adaptive behavior is the third problem: not whether a system can act, but when it should. Why timing is the hard part, why restraint reduces cognitive load, the challenge of shared state, and the trust condition — adaptation must be visible, reversible, and tunable. Boundary: trustworthy well-timed adaptation is active research. Links: https://varunsiddaraju.com/ · https://varuninnovates.com/ · https://veeruby.com/ #XR #AI #SpatialComputing #MixedReality #HumanComputerInteraction This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit varunsiddaraju.substack.com

    Adaptive Behavior — When Should AI Step In?

  6. Aug 4

    Spatial Memory — Why XR Keeps Forgetting You

    Note: This audio uses an AI-generated voice model of Varun Siddaraju’s voice, based on his original scripts. It is disclosed explicitly because this work is about trust, explainability, and human-centered systems. Spatial memory is the second of the four problems: holding a working memory of places, objects, tasks, and relationships across time — so you can leave and return without penalty. Why XR resets between sessions/devices/phases, why persistent selective memory is hard (and intimate), and why it’s becoming infrastructure, not a feature. Boundary: still mostly an open problem. Links: https://varunsiddaraju.com/ · https://varuninnovates.com/ · https://veeruby.com/ #XR #AI #SpatialComputing #MixedReality #HumanComputerInteraction This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit varunsiddaraju.substack.com

    Spatial Memory — Why XR Keeps Forgetting You

  7. Jul 21

    Perception — Reading Human State, Not Head Position

    Note: This audio uses an AI-generated voice model of Varun Siddaraju’s voice, based on his original scripts. It is disclosed explicitly because this work is about trust, explainability, and human-centered systems. Perception is the first of the four problems behind context-aware XR: reading continuous human state — attention, task, context — not just head pose. Why it’s an inference problem under uncertainty, why it gates well-timed assistance and cognitive-load reduction, and what good perception design looks like. Honest boundary: reliable human-state reading is still open research. Links: https://varunsiddaraju.com/ · https://varuninnovates.com/ · https://veeruby.com/ #XR #AI #SpatialComputing #MixedReality #HumanComputerInteraction This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit varunsiddaraju.substack.com

    Perception — Reading Human State, Not Head Position

  8. Jul 7

    Context Is the Real Interface

    Note: This audio uses an AI-generated voice model of Varun Siddaraju’s voice, based on his original scripts. It is disclosed explicitly because this work is about trust, explainability, and human-centered systems. Context-aware spatial systems — what they are, and why XR keeps failing the real world. Episode 1 lays out the thesis: persistence, context, and world-state matter more than interfaces. The four problems behind adaptive XR — perception, spatial memory, adaptive behavior, explainability — and why world models are becoming infrastructure. Honest boundary: this is active research, not a finished product. Links: https://varunsiddaraju.com/ · https://varuninnovates.com/ · https://veeruby.com/ #XR #AI #SpatialComputing #MixedReality #HumanComputerInteraction This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit varunsiddaraju.substack.com

    Context Is the Real Interface
  • 8 Episodes

About

Exploring the future of AI, AR/VR, and human-centered innovation—building bridges between technology, research, and real-world impact. varunsiddaraju.substack.com

Information