Jeremy Bailenson: Taking a grand tour of the latest in virtual reality

Episode Summary

Executive Summary: The episode explores virtual reality as a tool for reshaping perception, behavior, and learning, not just entertainment. Stanford researcher Jeremy Bailenson argues VR can induce real psychological effects despite imperfect fidelity, enabling applications in food research, health, education, empathy, and training. The conversation also highlights major technical barriers to social VR, especially accurate capture of nonverbal cues and facial expression.

Main Topics: What virtual reality really is (Priority: 5/5): Russ Altman and Jeremy Bailenson unpack the idea of 'virtual reality' as a computer-generated sensory environment that can approximate, but not reproduce, the richness of the physical world. They emphasize that VR works by creating an illusion that the brain treats as real. Behavior change through sensory manipulation (Priority: 5/5): Bailenson explains that VR can alter perception, self-image, and social behavior by controlling visual, auditory, haptic, and sometimes olfactory inputs. The goal is not perfect realism but reliable psychological influence. The virtual donut study and satiety (Priority: 5/5): A study tested whether virtual touch and smell around a donut could influence later eating behavior. Results suggested that virtual exposure increased satiation and led participants to eat fewer real donuts later, supporting a food-reduction strategy with possible health implications. VR in health research and emotion induction (Priority: 4/5): The lab uses VR to reliably induce emotions before MRI scanning, helping study depression, cognitive appraisal, and brain responses. VR is presented as a practical shortcut for creating emotionally rich lab conditions that are otherwise hard to reproduce. VR for children and social development (Priority: 4/5): Bailenson describes research on immersive VR with young children, including executive function, compliance, sharing, and reality blurring. He also notes VR’s promise for children with autism to practice social skills in low-risk simulated settings. Social VR, avatars, and the future of work (Priority: 5/5): The discussion turns to networked avatar systems as a possible replacement for some commuting and video conferencing. A key challenge is capturing subtle nonverbal behavior quickly enough to make remote interaction feel genuinely social. Commercial adoption and training as the killer app (Priority: 4/5): While entertainment gets the most attention, Bailenson argues that business training is the strongest current use case. He cites large-scale deployments at Walmart and Chipotle for stressful, high-stakes scenarios that are difficult to simulate in real life.

Key Arguments: VR is valuable not because it perfectly reproduces reality, but because it can create a convincing illusion that changes behavior and cognition. Sensory cues such as touch and smell strongly shape eating behavior; virtual exposure may reduce later consumption rather than simply prime cravings. VR can provide experimental control for emotion research, especially when paired with MRI and machine learning. Immersive VR may blur reality boundaries for young children, so developmental effects must be studied carefully. For social VR to feel natural, systems must capture nonverbal cues like gaze, posture, and facial micro-expressions with very low latency. Business training is currently the most compelling real-world VR use case, more so than consumer entertainment. Avatars may be more scalable than photorealistic video because they transmit movement semantics rather than full images, reducing network burden.

Data Points: Years of lab research: 20 years - Bailenson says his lab has studied VR simulations for about two decades. First donut study sample size: about 100 experimental participants - He describes the donut experiment as preliminary and small to medium in size. Virtual emotion study sample: north of 100 people - He references a VR emotion-induction study feeding into machine learning analysis. Child VR research sample: north of 200 kids - Jackie Bailey’s dissertation work in the lab involved preschool-aged children in VR. Children's ages in VR study: 3 to 6 years old - Young children interacted with Grover or Elmo in immersive VR studies. Large-scale employee training: over 200,000 employees - Walmart employees reportedly used VR for training in holiday-rush scenarios. Head-tracking rate: 30–40 times a second - The lab tracks head rotations in VR to infer emotion and engagement. Nonverbal timing: a tenth or a thirtieth of a second - Bailenson cites interactional synchrony and the speed of face-to-face nonverbal exchange. Car accident deaths referenced: 40,000 people - Used to underscore the potential safety benefit of reducing commuting.

Pivotal Quotes: "The illusion falls very far short of reality objectively, but for our brains, it's pretty real." — Jeremy Bailenson: He explains why VR can still be psychologically effective even when it is not fully realistic. "What if we could get rid of the commute?" — Jeremy Bailenson: He frames social VR as a potential replacement for some in-person travel and office routines. "The killer app that I've seen so far is training." — Jeremy Bailenson: He argues that enterprise training is the strongest current commercial use of VR.

Implications: VR’s near-term future is less about perfect worlds and more about targeted, measurable behavior change—especially in health, education, and training. The biggest hurdles are developmental safety, facial/gesture capture, and making avatars feel genuinely human.