Apple Vision Pro as a startup platform: opportunities and timing

Executive overview

Most AR/VR attempts failed because optics are a physics problem, not a compute problem. Apple sidestepped this by using a high-fidelity video passthrough and investing heavily in eye tracking, custom silicon, and spatial sensing — technology closer to a self-driving car than a phone.

The Vision Pro is positioned as a productivity device, not a gaming one. That's a meaningful departure that could make it something people use daily.

The core insight: Apple has done the hard platform work; the killer apps will come from third-party founders who figure out interactions Apple hasn't imagined yet — just like pull-to-refresh on iPhone.

What Apple actually built

• Pass-through video (fully digital feed) instead of optical AR — reduces optical physics constraints
• Foveated rendering: eye tracking determines where to render high-resolution pixels, saving power and heat
• R1 co-processor handles real-time sensor fusion from 10+ cameras, LiDAR, and IR depth cameras
• Same underlying tech as self-driving cars: SLAM (Simultaneous Localization and Mapping)
• Built on Apple's existing silicon, True Depth camera, and LiDAR expertise from iPhone and iPad

Why the SDK matters for developers

• visionOS apps are native SwiftUI: opening a PDF takes a few lines of code vs. many on Meta's platform
• Meta's SDK comes from a gaming DNA (Unity/Unreal); game engines treat the world as finite — real-world spatial computing doesn't fit that model
• Human Interface Guidelines for Vision Pro center on eye tracking and depth — a new design vocabulary to master
• Eye tracking UX is at the same inflection point capacitive touch was on the original iPhone

The iPhone analogy and timing

• iPhone launched without an app store; the breakout companies (Instacart, DoorDash, Coinbase, Uber) came 5–6 years later, when 70–80% of people had a device
• Vision Pro likely mirrors the iPhone launch moment, not the "category already mature" moment
• Mass adoption will take longer than mobile — the Vision Pro will likely capture high-end professional use first
• Tesla Roadster failure mode: success at the high end without a path to a mass-market follow-on

Advice for founders building on Vision Pro

• Don't wait — expertise compounds; being early builds a moat that's hard to fake
• The right founders are irrationally compelled to build VR apps in their spare time already
• First-principles evaluation beats platform thesis: ask whether the idea makes sense, not whether the platform is hot
• Focus on workflows where high information density and 3D space create genuine leverage (trading floors, CAD, engineering, construction)
• Third-party developers will define the new interaction paradigms — pinch-to-select is just the start