The state of spatial computing in 2023: Perspectives from two decades in the field

In this post, I share high-level impressions of key spatial computing trends in 2023. Topics covered are hardware, including the Apple Vision Pro and Humane Ai Pin; generative AI, including visual understanding and conversational input; and the metaverse, including VR adoption and divestiture. 2023’s developments indicate that it’s a good time to make sure you’re paying attention, learning relevant skills for spatial computing, and that you know where to go for help.

A colleague asked me for my impressions of the state of spatial computing at the end of 2023, which prompted me to write them up for easier sharing.

I’ve been researching spatial computing interfaces since 2005, and the industry has changed dramatically in that time.  In this post, I’ll share my high-level impressions of key spatial computing trends in 2023.  This is not a formal analysis, but it is formed by my nearly two decades of experience in the field.

As a customer strategy consultant, my focus is not on any single device or technology, but on understanding how users will embrace and respond to spatially embedded interfaces over time.  I’m most interested in recognizing how the public conceptualizes spatial computing and evaluating how real human needs will form the spatial computing businesses of the future.

In this post, I’ll cover three major trends:

  • Hardware, including the Apple Vision Pro and Humane Ai Pin

  • Generative AI, including visual understanding and conversational input

  • The metaverse, including VR adoption and divestiture

Hardware

The Apple Vision Pro is one of two devices that caught the attention of even non-spatial computing circles.

The Vision Pro announcement was impressive, but it displayed the same issue everyone is grappling with in augmented reality: there was no killer use case.  Apple’s demos retreaded old ground, but they did so beautifully.  But with Apple releasing a developer kit, there is now an army of enthusiasts trying to suss out viable use cases.

The more subtle thing Apple has done, which I think will be invaluable if it sticks, is to make the umbrella term of “spatial computing” more visible to a broader audience.

Up to now, different branches of spatial computing technologies have evolved independently.  There are vision AI circles, which are separate from IoT and smart device circles, which are separate from AR/VR/XR/MR circles, plus dozens of adjacent technologies such as robotics which all have their own separate communities, cultures, and processes.  Yet from a human-centered design perspective, all of these communities are solving overlapping problems.

“Spatial computing” is the first umbrella term to achieve any sort of mainstream adoption.  I anticipate that simply having overlapping language will begin to create bridges between technology silos and open the door to more effective cross-pollination and human-centered design approaches.

The other device that caught the attention of non-spatial computing circles in 2023 was the Humane Ai Pin.

In late 2023, Humane announced a device that attaches to your lapel and sees what you see.  You can query it about the scene in front of you, and it can project an interface onto your hand.

What I find most interesting about Humane’s Ai Pin is its form factor.  Humane is solving augmented reality-type problems without forcing users to wear anything on their face.

Head-mounted displays have long been a stumbling block for adoption of augmented reality.  Having your face covered, even partially, is off-putting for people in social interaction, and repeatedly pulling a helmet off and on has always been a friction point even for developers who spend their entire day creating immersive experiences.

That Humane found a way to provide similar value without the head-mounted display is further evidence that spatial computing is the future of the tech industry.  But it may not look exactly as we expect, and it may not be mediated by the devices we recognize as “spatial computing devices” today.

Generative AI

At first glance, generative AI doesn’t appear to have a lot to do with spatial computing.  But as evidenced by the Ai Pin, these two technologies will inevitably grow together with time.

Generative AI is now at a point where you can upload a picture, and it can “understand” it well enough to realistically modify elements within that scene, keeping the same style, perspective, and lighting.  It can also make sense of things like content and composition, which means it can generate similar images in a different style, or with different figures and components; and with varying success, it can do all this in response to language-based prompts.

Even acknowledging that AI “understanding” of scenes is more like autocomplete than human comprehension, what this amounts to is a new kind of environmental sensor that developers can start to incorporate into spatial computing devices.  Uploading digital images is only a step away from providing AI with real-time images from a front-facing camera (something that’s already been explored with projects like Microsoft’s Seeing AI).

Generative AI–and especially ChatGPT–has also dramatically increased consumer expectations for conversational interfaces, making the voice assistants of the 2010s feel rather dated.  (And Amazon has responded: they announced in September that they will be incorporating ChatGPT-like features into their Alexa devices–not a minute too soon.)

Voice is one of the most natural inputs for spatial computing devices.  Better comprehension of voice commands will dramatically increase the efficiency with which users can control objects and devices without traditional screens.

However, this increased controllability of smart devices comes with associated risks.  2023 has shown how misplaced trust in ChatGPT can go wrong in journalism, healthcare, and even the courtroom.  How much worse will these issues get when generative AI has free reign over our environments?  This will be an important question for spatial computing designers to grapple with as AI and spatial computing technologies become more integrated.

 
 

Virtual reality and the metaverse

Other indicators of where spatial computing is headed come from looking at where we’ve already been, particularly virtual reality and the metaverse.

Virtual reality is down but not out.  Despite reports that device sales dropped in 2023, VR continues to have a passionate userbase among gamers and younger audiences.  The initial buzz has dwindled, but core users are still finding unique value in what VR offers nearly seven years after the Oculus Rift was released.

This is different from failed technologies like 3D TV (which fizzled out more quickly and was less accessible to younger audiences).

This suggests that immersive technology isn’t going anywhere, and that consumer expectations for where and how this technology can be used will continue to rise.  I also anticipate there will be some generational effects at play in the long term, as younger audiences age up and become dissatisfied with the limitations of traditional flat-screen-only interfaces.

2023 was also the year of tech layoffs (which unfortunately are continuing in 2024), and press reports indicate that teams working on augmented reality and metaverse technologies have been particularly hard-hit.

It’s been painful to see these teams and projects gutted.  It also wasn’t without precedent.

Anyone who’s worked in innovation long enough knows they’re the first on the chopping block in an economic downturn.  And as evidenced by the Vision Pro (discussed earlier), augmented reality especially is still struggling to find high-value use cases that will drive robust sales.

But I think the chickens were also coming home to roost for the Zuckerbergian vision of the “metaverse,” not only for Meta, but for competitors who reactively scrambled to assemble their own social VR teams.

I do believe there are viable use cases for both social VR and connected spaces.  But the Horizon Worlds strategy, as initially concepted, overlooked fundamental human motivations required for long-term adoption of social VR.  (This criticism is in no way meant to disparage the many talented people who have worked on Horizon Worlds.)

In the coming year, I expect immersive technology to be treated by big tech companies more like the emerging technology it is, with smaller budgets, but also less immediate pressure to deliver.  Spatial computing innovation teams will need to run leaner shops.  This means making every investment count through a culture of entrepreneurially focused research (which is a place where I can provide support).  Teams that experiment more often and more effectively will pivot more aggressively and find bigger wins faster.

Gradually, then suddenly: edging closer to a spatial computing sea change

In this post, I’ve covered three major trends:

  • Hardware, including Apple Vision Pro and Humane Ai Pin

  • Generative AI, including visual understanding and conversational input

  • The metaverse, including VR adoption and divestiture

Here are some closing thoughts.

In 2023 (and as of early 2024), spatial computing is still flying under the radar.  The industry has not yet reached its tipping point.  As a consultant, I don’t have companies crawling out of the woodwork with requests for spatial computing projects.

But at the same time, key developments in 2023 mean that we are edging ever closer to a sea change.

It would be dangerous to underestimate the attention Apple’s Vision Pro (releasing in February 2024) may bring to spatial computing, both from the perspective of use cases and terminology.

Interest in spatial computing via virtual reality devices also persists despite challenges, while new technologies offer a strong glimpse of previously unanticipated possibilities: new form factors via Humane’s Ai Pin (and Humane’s competitors to come), as well as rapid evolution of advanced input capabilities via generative AI.

Corporate investments in spatial computing may have slowed in 2023, but this offers teams the opportunity to experiment more boldly with these new developments out of the limelight.

Management scholar Rita McGrath says that strategic inflection points in business occur “gradually, then suddenly.”  When they do occur, they redefine the basic assumptions of business, toppling established leaders and giving smaller players a rare opportunity to become leaders themselves.

McGrath’s strategic inflection point curve

In 2023, spatial computing is still in its “gradually” phase, but there are warning signs that “suddenly” is around the corner.

For businesses, it’s a good time to make sure you aren’t giving up on your investments in spatial computing, however slow things may seem in light of 2023’s layoffs.

And for individuals, it’s a good time to make sure you’re paying attention, learning relevant skills for spatial computing, and that you know where to go for help (something I’m always happy to assist with, whether it’s me or pointing you to one of my colleagues).

It's an exciting time to be working in spatial computing, and I’m looking forward to seeing where things go in 2024.

If you’re looking to bring an evidence-based, human-centered lens to an upcoming spatial computing project, I can help. You can schedule an introductory conversation with me here.


References and further reading

Generative AI

Generative AI headlines in 2023

Entrepreneurial experimentation

Strategic inflection points

Llewyn Painespatial computing, vision pro, augmented reality, ai pin, generative ai, conversational interfaces, voice assistants, virtual reality, metaverse