FEATURE
ENGINEERED TO REDEFINE REALITY
From real-time transcription to AI-powered personal assistants, the convergence of AI and extended reality is reshaping accessibility, consumer wearables, and the future of the human-machine interface.
Written by Poornima Apte
Illustrations by Gabriel Corbera
PICTURE DINING WITH YOUR FRIENDS in a noisy restaurant where you can barely hear what they are saying. What if, instead of straining hard to pick up the conversation, you could read a transcript of it in your eyeglasses? The idea underscores XanderGlasses, glasses that transcribe conversations in real time so people with hearing loss can read.
When XanderGlasses co-founder Alex Westner came across the smartphone game Pokémon Go, where users catch and train Pokémon characters, he suspected there could be a more meaningful use for augmented reality (AR). A graduate of MIT’s Media Lab, Westner had spent his career trying to help computers understand sound. Through the XanderGlasses, the technologies—sound, AR, and AI—work together to help people with disabilities.
The glasses highlight the malleability of technologies like AR, VR, MR, all of which fall under the XR umbrella, to adapt to practically any medium or form factor demanded by the next wave of advancements. For a while, smartphones were the medium of choice for consumers but form factors like eyeglasses, pendants, rings, pins, and more are vying with each other as the next smart device.
DEFINITIONS
Virtual Reality (VR)
Immerses the user in a completely synthetic, computer-generated environment that mirrors but replaces the physical world.
Augmented Reality (AR)
Overlays digital information—such as graphics, data, or contextual cues—onto the user’s view of the physical world in real time.
Mixed Reality (MR)
Blends physical and digital worlds so that virtual objects coexist with the real environment and can interact with it in a spatially aware manner.
Extended Reality (XR)
An umbrella term that includes AR, VR, and MR, representing the gamut of technologies that blend physical and digital environments.
Video: XanderGlasses
Technology Advances
Getting to the Goldilocks of consumer wearables—feels just right in form and function—is not easy and depends on many technologies advancing in parallel. Indeed, running compute-hungry algorithms on small devices might sound fairly run-of-the-mill today but it took a lot of systematic technology improvements along the way for us to get here, said Alex Katouzian, executive vice president and group general manager of Mobile, Compute, & XR (MCX) at Qualcomm Technologies.
Take XanderGlasses as an example. They don’t rely on the internet or the cloud; instead, all processing components, including microphones to capture audio, are built in. “It’s like having a low-end smartphone on your head,” Westner said. The device converts received audio from speech to text through a large vocabulary AI model on the unit. “It’s understanding syllables, words, and sentences and making sense of the sounds to create meaningful sentences with punctuation,” Westner said. The text renders onto the display screen at about three feet in front of the eyeglasses for reading ease.
While many technologies had to come to fruition for XanderGlasses to move from drawing board to reality, Westner credits two primary ones: speech-to-text AI and mini hardware. “The technology finally got to the point where the projectors (for text display) were small and efficient enough to be able to miniaturize the optical system and the battery required to power them was not so big that you needed to wear a helmet,” Westner said.
“It was the efficiency of the optical system that allowed AR glasses to become more acceptably stylish,” Westner added, “otherwise no one’s going to wear goofy-looking things outside. Especially when you already have a stigma from having hearing loss, you don’t want to draw attention to yourself.”
Miniaturization and integrating functionality into increasingly smaller packages are part of the progress toward exploring new form factors for XR. Other significant technology improvements include an embedded neural processing unit (NPU) that doesn’t drain battery life, video and camera capabilities, and data security, Katouzian pointed out.
But above them all, one technology is improving the utility of XR to the point where it’s poised to become mainstream: AI.
The Intersection of AI and XR
We’re at a “really fun moment where the XR industry can think more expansively with AI built in,” said Elizabeth Hyman, CEO at XR Association, a trade organization that promotes responsible development and thoughtful advancement of XR.
It’s noteworthy that XR and AI have always been mutually symbiotic, Hyman said. “People are going to be looking for human ways to interact with artificial intelligence and XR is poised to deliver that in a way that feels more like a human interaction,” she said. In turn, AI has helped develop the world-building features of XR, Hyman added.
George Lecakes, director of Immersive Technology and Learning at Rowan University, agreed. “AI has helped create environments in the virtual world, those that would be otherwise notoriously time-consuming and expensive,” said Lecakes, who also serves as associate director of virtual reality at Rowan’s Machine and Artificial Intelligence VR Center (MAVRC). “We can use AI systems to generate at least the ground floor of a virtual world and we can then have artists tweak these so development is a lot faster.”
AI is also shaping VR by personalizing instruction, said Nidhal Bouaynaya, a professor and associate vice president for artificial intelligence at Rowan University, and director of MAVRC. Through Rowan’s Dreamscape Lab, medical students can not only learn procedures through immersive training scenarios but can also access personalized review protocols through embedded AI, so every student gets the most out of all training. In turn, VR can help AI by filling gaps in data sets in training models. Such synthetic data has helped complete edge use cases where real-life data might not be readily available.
Over the years, Hyman has been tracking the steady growth of XR, especially its use in enterprise workforce training and development. For example, an immersive AR or VR headset can help a new employee on a manufacturing floor learn the safe operation of a die cast machine before they get their hands on a real one. “Embedding AI in smart glasses adds context,” Hyman said, “it gives the user the contextual environment and delivers information they need without having to consult a manual.”
In 2025, Rowan University opened its Dreamscape Learn Center. Outfitted with cutting-edge VR headsets, sensors, and tactile feedback tools like joysticks, fans, and vibration systems, two dedicated VR classrooms allow students to collaborate or work individually inside interactive worlds. Photo: Rowan University
A close-up view of the XanderGlasses. Photo: XanderGlasses
AI-XR Consumer Wearables
That same training support on the manufacturing floor translates over to personal assistance for consumers through smart glasses, especially because the greater public is now more familiar with the concept. With AI-embedded smart glasses, you can access directions, use them as a teleprompter, review appointments on a calendar, and yes, even read captions for a conversation, without ever having to consult your phone.
Essentially, the XR-AI combo is accelerating a new age of personal assistants. The 2026 edition of the Consumer Electronics Show (CES) was all in on personal AI within consumer device ecosystems, reported Ben Chan, industry analyst for technology intelligence firm, ABI Research in a post-CES roundup.
The focus is on integrating multiple data points from devices ranging from mainstream smart devices like smartphones, laptops, and smart watches to new form factors like AR-enabled smart glasses and perceptive wearables into a centralized AI agent for daily living.
Understanding the potential of banking on such an integrated ecosystem of devices, the large technology companies like Google, Apple, Amazon, Meta, and the like are all fighting to be the assistant of choice for the consumer, Katouzian said. The field is wide open.
Challenges to Overcome
For now, despite the many advances in XR, especially because of its integration with AI into small form factors, the focus is on delivering hardware that is power-efficient. “We need better battery life and lighter modalities,” Hyman said, “you have to ensure the device will maintain its comfort level, from heat and other things.”
The XR Association reports that key barriers to integrating XR, AI, edge computing, and advanced connectivity in advanced manufacturing include: high upfront costs, complex integration with legacy systems, fragmented or poor-quality data, significant workforce skills gap, workforce hesitancy to new technologies, cybersecurity concerns, limited supply of relevant XR applications, and a difficulty demonstrating a clear return on investment—especially for smaller firms.
“Overcoming these obstacles will require coordinated efforts in standards development, government adoption and incentives, robust training programs, and large-scale demonstration projects to build confidence and accelerate adoption across the manufacturing sector,” the association offered as part of its input to the Office of Science and Technology Policy (OSTP) for the development of the National Strategic Plan for Advanced Manufacturing.
U.S. legislation has been introduced to spur advances in XR. One is the United States Leadership in Immersive Technology Act of 2025. It establishes an Immersive Technology Advisory Panel to promote the use of immersive technology in the United States. There’s also the Immersive Technology for the American Workforce Act of 2023, a competitive grant program to support community colleges and career and technical education centers in developing immersive technology education and training programs for workforce development, and other purposes.
Photo: Getty Images
Photo: Getty Images
Back to the Future
Advances in XR are coming fast and furious, many of which seek to redefine what the future of the human-machine interface might look like.
Accessing AI interfaces through smart glasses or other small devices is not seamless and device vendors are working to ease these friction points.
What will the human-machine interface of the future look like? There might be early clues in Apple’s late January 2026 acquisition of Q.ai, a startup that interprets tiny facial movements or near-silent speech.
“For Apple, this could become a key differentiator in the race for consumer AI glasses and wearables, enabling communication with an AI assistant through a whisper or barely perceptible mouth movement,” said Helen Papagiannis, an expert on AR and spatial computing.
“If Apple can make silent speech reliable, it becomes foundational, the way multitouch reshaped the iPhone or Face ID redefined authentication,” Papagiannis said. In essence, we will be seeing a new class of interface that feels like a natural extension of the human mind.
Poornima Apte is a technology writer based in Walpole, Mass.
© 2026 The American Society of Mechanical Engineers. All rights reserved.