Sequoia's Shaun Maguire: hardware is coming back, AI video gets its ChatGPT moment within 12 months, robotics in 3 years

Summary

Shaun Maguire, partner at Sequoia Capital, argues that hardware is not returning as a trend but reasserting itself as the foundational layer of the current technology cycle, a view he first circulated internally at Sequoia roughly three and a half years ago in what he describes as a hardware manifesto shared with approximately 100 people across Silicon Valley.

The core thesis draws on Sequoia's own history. For its first 25 years, the firm made almost all of its returns in hardware. The following 25 years were dominated by software. Maguire's forecast is that the next 25 years revert to hardware, with AI serving as the catalyst.

The Digital-Physical Gap

Maguire frames the current moment around a ten-year gap between digital and physical intelligence deployment. Digital intelligence, in his view, reaches abundance roughly a decade ahead of physical intelligence, not in peak capability but in diffusion rate and full economic rollout. That gap creates a sustained window where hardware remains uncommodified while software increasingly faces margin compression.

He sees Moore's Law-era silicon as a branching technology tree, with the next phase extending into humanoid robotics, silicon photonics, and orbital data centers, each representing decades of compounding progress and the potential for large standalone businesses, alongside significant capital losses.

AI Video Before Robotics

On timing, Maguire offers two concrete forecasts. He expects a ChatGPT-equivalent consumer breakout moment for AI video within the next 12 months, characterising the current state as analogous to the GPT-3 era, where model quality is recognised by insiders but the consumer packaging has not yet triggered mass adoption.

For robotics, he sets the public inflection point at roughly three years out. He distinguishes between industrial utility, where robots like Optimus will be performing meaningful work inside factories sooner, and the broader consumer moment where the technology feels ubiquitous and personally transformative. The latter, in his view, lags private deployment by several years.

Reframing the Pivot Narrative

Maguire pushes back on the framing that energy or industrial companies are pivoting into AI. His inversion is that AI is becoming an energy industry, and that Silicon Valley is belatedly recognising the scale and sophistication of supply chains, chemicals, and semiconductor manufacturing that predate and underpin everything the Valley has built. He points to former Bitcoin miners such as Crusoe and CoolWeaver as early examples of capital rotating toward AI infrastructure, but frames that as convergence rather than pivot.

Lunar Manufacturing

On Elon Musk's mass driver concept for lunar manufacturing, Maguire's position is that the hardest prerequisites, Starship and Optimus, are nearly in hand, and that the actual engineering of a mass driver and lunar fab is comparatively straightforward given existing terrestrial supply chain knowledge. His timeline is 15 to 25 years, with the key constraint being time and execution rather than fundamental technical barriers.