Shaun Maguire on the Elon empire, Tesla Optimus robots, and his bet on Reflect Orbital

Summary

Shaun Maguire, partner at Sequoia Capital, lays out his thesis on humanoid robots, space infrastructure, and why he backed Reflect Orbital — a startup with an idea most investors would dismiss on first hearing.

Tesla Optimus

Maguire attended the Tesla Optimus demo in Los Angeles and says the in-person experience was meaningfully more impressive than video coverage suggested. Twenty untethered robots walked out from behind a curtain and interacted with the crowd for roughly two hours. He says he initially couldn't tell at twenty feet whether he was looking at a human in a costume or a machine — the tell was the narrowness of the waist. His read is that most of the actuation was genuinely autonomous; a human operator may have provided high-level goals, but the low-level motor control and feedback loops were running independently.

He frames humanoid robots as the physical-world corollary to foundation models: if AGI handles unlimited knowledge work, robots handle unlimited physical work. He's confident the category is real — the only open question is timing, which he says is genuinely hard to forecast given supply chain complexity across actuators, sensors, rare earth magnets, and manufacturing scale.

Sequoia is not an investor in Tesla, but Maguire notes the firm holds major positions in SpaceX, Boring Company, X, and xAI.

Reflect Orbital

Reflect Orbital's plan is to deploy a constellation of satellites that orbit permanently along the day-night boundary and reflect sunlight back to specific ground targets. The near-term business is lighting — think wildfire suppression at night, disaster relief, or extended construction hours at high latitudes. The longer-term business is energy.

The energy thesis turns on a straightforward economic lever: a solar farm generating power for roughly ten hours a day carries years of debt repayment before it becomes profitable. If Reflect can extend effective generation by around four hours — two before sunrise, two after sunset — that incremental output is almost pure profit for the farm operator, because the silicon panels are already on the ground with no additional operating cost. Reflect's model is a revenue share on that incremental energy. Maguire's framing is that the company requires no ground infrastructure investment; the ~10,000 solar farms already built globally are the distribution network.

He argues the business couldn't have existed ten years ago even if launch costs had been low enough, because there weren't enough solar farms to sell into. Both conditions — cheap launches and ubiquitous solar infrastructure — only now exist simultaneously.

The SpaceX dependency question

Maguire is direct about why he's comfortable building around SpaceX's cost curve rather than treating it as a threat. He expects Starship to drive launch costs down dramatically over the next five years and sees mass-to-orbit capacity already rising sharply. His view is that investors who wait for those cost reductions to materialize before backing space companies will find themselves late and in competitive markets. The time to build is now, so a company can be scaling when the curve inflects.

On the risk of SpaceX entering adjacent markets, he points to regulatory exposure and Elon Musk's focus on Mars and Starlink as structural buffers. He thinks SpaceX is incentivized to foster a broad ecosystem of independent companies in low Earth orbit rather than compete with them.

The Varda miss

Maguire volunteers that he passed on Varda — Delian Asparouhov's space manufacturing company — at the seed stage and calls it a mistake. His reasoning at the time was that Varda's first product, ultra-pure fiber optic cable for long-range communications, would be made irrelevant by Starlink, which he believes has a 10x-plus cost advantage per gigabyte over undersea fiber for long-haul data. He still holds that view on the product, but says he was wrong to let a skeptical micro thesis override a correct macro thesis: that the timing for space companies was right and that a strong team of space engineers would find a valuable product even if the initial one was wrong.

Founder advice

Maguire's advice for new graduates is blunt: don't start a company without a genuinely clear idea. Join SpaceX, X, Varda, or another high-quality organization for a year or two first. When you do have an idea, ignore what investors — including him — tell you, because the best ideas routinely sound crazy at the outset.