$6.5 Billion. Still not ready.

What OpenAI's Delays Reveal About the Real Cost of Industrial Design

In May 2025, OpenAI acquired io Products, the hardware startup co-founded by legendary designer Jony Ive, in an all-stock deal valued at approximately $6.5 billion. Not for code. Not for a patent portfolio. Largely for Ive's design judgment and the team assembled around him. mexc

That number is worth sitting with — and so is what's happened since.

The device's timeline has shifted repeatedly since the acquisition was announced. OpenAI's chief global affairs officer confirmed at Davos in January 2026 that the company remained "on track" for a second half 2026 launch, but reporting throughout the year has continued to push expectations later. Internally, there have reportedly been tensions between OpenAI's existing hardware team and Ive's design team over the speed and secrecy of design revisions, with Ive making the final call on almost all design decisions. Design News + 2

For anyone working in industrial design in Sydney or anywhere else, none of this is surprising. It's the most familiar tension in the discipline, just playing out at a scale most of us will never operate at.

Why Design Takes the Time It Takes

Design for manufacturing is not a linear process. It cannot be compressed into a sprint the way engineering milestones often can. Understanding a user, observing real behaviour, prototyping, testing, and refining — this is work that resists artificial deadlines because the insights it produces simply aren't available on demand.

Engineering has a clear failure signal. If the code doesn't compile, if the circuit doesn't power on, everyone in the room knows immediately. The problem announces itself.

Industrial design's failure signal is almost invisible by comparison. Skip or rush the design process and the product still gets manufactured. It still ships. Nothing visibly stops the assembly line. The cost shows up later — in a feature nobody quite understood, an interaction that feels needlessly difficult, a product that functions perfectly but that nobody particularly loves.

This asymmetry is the single biggest challenge facing industrial design as a discipline. The value of doing it well is hard to point to directly, because the alternative — doing it badly or not at all — doesn't produce an obvious failure. It produces a quiet underperformance that competitors rarely get credited for noticing, and that the affected company often never properly diagnoses.

What OpenAI's Experience Reveals

OpenAI is arguably the best resourced technology company in the world attempting to solve this exact problem. Unlimited capital. Some of the best engineering talent available. And still, by their own account, the constraint that keeps moving their hardware timeline is design — not chips, not manufacturing capacity, not supply chain.

That should tell every founder pursuing hardware product development something important. If a company with $6.5 billion to spend on design leadership alone still can't compress the design process, smaller hardware startups attempting the same shortcut are unlikely to fare any better.

The lesson isn't that design is slow for its own sake. It's that genuinely good industrial design requires depth of engagement with a problem that cannot be rushed without the product showing it later.

The Real Cost of Skipping It

You can buy the talent, as OpenAI has done. You can assemble the best design team money can find. But you cannot shortcut the process that team needs to do its work properly — and the market rarely punishes you immediately for trying. It punishes you quietly, later, in outcomes that are hard to trace back to their cause.

This is precisely why industrial design in Australia and globally remains undervalued relative to its actual impact on a product's success. The discipline's biggest obstacle has never been proving that good design matters. It's that the cost of skipping it doesn't show up where anyone is looking for it.

Even Jony Ive, working with effectively unlimited resources, is finding that out in real time.

This post is part of an ongoing series on why depth of engagement in industrial design cannot be automated, outsourced or skipped — and what happens to products when it is.

Danny Cheung is the founder of Paranormal Design, a Sydney-based industrial design consultancy specialising in hardware product development and design for manufacturing.