Virtual Reality: A History of Premature Arrivals, and What the Evidence Actually Supports

EveryEvery wave of virtual reality coverage, going back to the 1960s, has opened with some version of the same sentence: this time, VR is finally here. It was said about a fairground-style arcade machine in the early 1960s, about a military lab's ceiling-mounted rig in 1968, about a glove-and-goggles company in the mid-1980s, about a games console in 1995, about a Kickstarter campaign in 2012, and about a headset in 2024. In the narrow sense that the hardware worked and shipped, several of those arrivals were real; none, though, was the arrival the coverage promised. The useful question by now isn't whether VR has finally arrived — it's which specific capability actually got adopted each time, under what name, and what the evidence says about the handful of uses people keep claiming for it next.

Before the Headset Had a Name

MortonMorton Heilig, a cinematographer, spent the 1950s developing what he called "Experience Theater," laid out in a 1955 paper of the same instinct under a different name, "The Cinema of the Future." ¹ The machine he eventually patented, the Sensorama, granted in 1962, was a single-user arcade cabinet combining a stereoscopic 3D viewer, stereo sound, vibration, wind, and smell: its five short films included a motorcycle ride through Brooklyn, complete with engine vibration and city smells timed to the picture. With no head tracking and no computer behind it at all, every sensation was pre-recorded and fixed. Heilig couldn't find financial backing to put it into production, and the Sensorama never shipped as a commercial product.

What gets skipped in most tellings is that Heilig also patented the first actual head-mounted display, the Telesphere Mask, around the same period — eight years before the device usually credited with that title. The Telesphere Mask showed stereoscopic images and stereo sound to a user wearing it, but like the Sensorama, it played back fixed photographic or film content rather than anything generated or tracked by a computer. That distinction is the one that actually matters for what came next.Heilig's second patent beat Sutherland by eight years

The First Computer-Driven Head-Mounted Display

IvanIvan Sutherland built the device usually credited as the first real virtual reality headset in 1968, at Harvard and then the University of Utah, with graduate student Bob Sproull. Nicknamed the "Sword of Damocles" because its two cathode-ray tubes had to be suspended from the ceiling, too heavy to wear unsupported, it tracked the wearer's head and updated a simple wireframe room in real time to match ². The graphics were a handful of glowing lines, nothing like an immersive scene by any later standard, and funding for the work reportedly came in part from the CIA and Bell Helicopter rather than from an entertainment budget. What made it the first is exactly the thing the Sensorama and Telesphere Mask didn't have: a computer generating the image fresh, in response to where the user's head actually was, rather than playing back something recorded in advance.

A Name, a Company, and a Bankruptcy

TheThe phrase "virtual reality" itself is usually credited to Jaron Lanier, though sources differ on whether he coined it outright or popularised a term already circulating. What's firmer is the company: Lanier and Thomas Zimmerman left Atari in the mid-1980s to found VPL Research, the first company to sell VR goggles and a wired data glove as commercial products, refining the ideas Sutherland had demonstrated two decades earlier into something a customer could actually buy ³. VPL filed for bankruptcy in 1990; Sun Microsystems bought its VR and graphics patents in 1999. Nintendo's Virtual Boy launched in 1995 with a display running at 384×224 pixels in red monochrome and was discontinued within a year. The same year, Forte Technologies released the VFX1 Headgear, a PC peripheral aimed at serious gamers, at a US launch price of $695 — nearer $1,500 in 2024 money, and reported considerably higher again through resellers outside the US ⁴. An expensive, enthusiast-only price tag for "real" VR hardware is not something 2024 invented. Gartner's first-ever Hype Cycle chart, also published in 1995, already placed VR in the Trough of Disillusionment — a technology that had already peaked and crashed before the framework existed to describe the pattern ⁵.

Two More Waves, Tracked in Real Time

TheThe next wave is better documented than any before it, because Gartner was by then tracking it year over year rather than describing it after the fact: the Oculus Rift's 2012 Kickstarter campaign, Facebook's $2 billion acquisition of Oculus in 2014, and the near-simultaneous 2016 consumer launches of the Rift, the HTC Vive, and Sony's PlayStation VR. Gartner's own tracking put this second wave on the Slope of Enlightenment by 2017 and removed VR from its chart entirely by 2018, officially because the underlying display, tracking, and latency technology was judged mature enough to no longer count as emerging — the fuller account of that specific climb, and why it counts as recycling rather than simple recovery, is developed in this site's own hype-cycle case-study page rather than repeated here. What followed next is the part that case study doesn't cover: Meta's 2021 rebrand around the "metaverse," and the gap between that pitch and what actually happened. Horizon Worlds reached only about half its projected 2022 user numbers, venture capital funding for AR/VR startups fell by more than 90% between 2022 and 2023, and Meta's own Quest headset sales were down 16% year-on-year as of November 2024 even as display resolution kept climbing, from the Virtual Boy's 384×224 pixels in 1995 to the Vision Pro's 3800×3000 at its 2024 launch ⁶.

What 2025 Actually Shipped

TheThe newest data changes the shape of the story again, and in a direction none of the 2021-era pitches anticipated. IDC's year-end figures for 2025 show the overall XR device market growing 44.4% — but that growth was carried entirely by smart glasses without any display at all, which now account for the majority of XR shipments, while traditional VR and mixed-reality headset shipments kept declining. Even within Meta's own 72.2% share of the headset market, Quest lineup shipments fell 42.3% year-on-year in 2025, with Ray-Ban Meta glasses, not the Quest, described as the primary driver of the category's overall growth ⁷. Apple's Vision Pro, after an M5 chip refresh in 2025, reportedly sold around 390,000 units for the year, an estimated 5.2% of XR value rather than units ⁸. Built on Google's Android XR platform, Samsung's Galaxy XR launched in 2026 as the closest Android-based rival to the Vision Pro. The part of "virtual reality" that is actually growing right now is the part with a camera and no display, not the part with a headset and a virtual world inside it.growth, but not in anything with a screen

Grading the Future-Applications Claims Against the Evidence

ListsLists of VR's coming uses tend to read the same way regardless of decade: education, medicine, entertainment, remote work. Three of those claims have real evidence behind them by now, graded below by how strong it actually is.

• Education. A 2022 meta-analysis of 48 studies and 3,179 students found immersive VR produces a small overall positive effect on learning outcomes (Hedges' g = 0.38), with a stronger effect in K-12 settings (g = 0.61) than in higher education (g = 0.31) ⁹. That's a genuine, measurable benefit, not a transformation of how learning works — closer to a better textbook than a different kind of classroom.
• Exposure therapy. This is VR's best-evidenced application by a clear margin. A 2026 meta-analysis of 26 randomised controlled trials covering 1,649 participants found VR-based exposure therapy reduced specific-phobia symptoms with a large effect (Hedges' g = −0.98) and PTSD symptoms with a moderate one (g = −0.51) ¹⁰. Earlier reviews specific to phobias found VR exposure performed about as well as real-world in-vivo exposure, which is the harder comparison to beat.
• Surgical training. The evidence here is mixed rather than settled: a systematic review of orthopaedic trials found six of eight randomised comparisons favoured head-mounted VR training over traditional methods, while reviews in other surgical specialties report inconsistent results between studies, and almost none of the existing trials measure patient-level outcomes rather than trainee performance scores.

Entertainment and remote-work applications — 360-degree film, virtual concerts, virtual offices — are real products today, but none has produced anything close to the controlled-trial evidence base the three uses above do; treat claims about their future growth as plausible rather than demonstrated until that changes.

The Pattern Worth Noticing

AcrossAcross six decades, two things have stayed true at every stage: the underlying hardware kept getting better on every measurable axis — resolution, tracking latency, field of view, weight, cost — and the specific product that actually found users kept changing shape. An arcade novelty became a classified military demonstrator, became a bankrupt glove-and-goggle startup, became a failed games console, became a crowdfunded gaming peripheral, became a corporate metaverse pitch that missed its own targets, and is, as of 2025, mostly succeeding in the form of camera glasses with no display in them at all. None of that means the technology was a dead end at any point along the way, since the compounding hardware improvement is real and cumulative, but it does mean that "is VR finally arriving" has been the wrong question every single time it's been asked, because the thing that actually arrives is never quite the thing the previous wave's pitch described. The better question, every time, is which specific capability is being adopted under which name this time, and whether the evidence for the use case in front of you looks like the exposure-therapy numbers above or like the 2021 metaverse projections.

Related Topics

• Building for Immersive Platforms: Engine Choice and the Risks the Engine Doesn't Handle — what to actually do differently if you're building on this technology today, including the data-protection and market-reality questions this page's history sets up.
• Reading Seven Technologies Through Both Models — the detailed, Gartner-tracked account of VR's second wave as a hype-cycle case study, which this page builds its own account of 2012–2018 on top of.
• Being the God of a Virtual World: Bartle's "How to Be a God" — what it actually means to run a persistent virtual world once the hardware to enter one is cheap enough that people do.

References