Figure AI's Figure 02 has been performing real production work — inserting sheet metal components, sorting parts, transporting materials — at a BMW plant in Spartanburg since mid-2025. This isn't a demo. It's a working pilot inside an actual auto factory.

Confirmed

Cumulative investment in humanoid robotics companies had exceeded $6 billion through Q1 2026, making it one of the most capital-intensive emerging technology categories tracked. Figure AI has raised more than $1.1 billion, backed by Microsoft, OpenAI, Nvidia, and Bezos Expeditions; Apptronik has raised over $550 million with Google and Mercedes-Benz among its backers; 1X Technologies has raised over $225 million. Figure is reportedly targeting production of 1,000-plus units in 2026, scaling toward 10,000-plus in 2027 — if achieved, this would make Figure the first humanoid manufacturer to reach meaningful production scale. Tesla's Optimus Gen 2 is currently deployed only within Tesla's own facilities, performing tasks in battery assembly and logistics, with no confirmed external customer having received a unit as of April 2026.

The market is converging on a $40,000-60,000 price point for commercial-grade humanoid hardware, with Robot-as-a-Service subscription pricing expected to bring effective monthly costs down to $2,000-5,000. A widely cited 2023 McKinsey analysis of automation economics projected that roughly 30% of tasks across US occupations could be automated by the mid-2030s using currently available or near-available technology — a figure describing task automation potential broadly, not specifically humanoid robots, and predating the current wave of humanoid deployment.

Genuinely uncertain

Industry analysis explicitly frames 2026 as a year where "humanoid robots supplement rather than replace human workers" — the current deployments are real but narrow: specific repetitive tasks in controlled factory environments, not general workplace replacement. Whether that changes as the technology matures, and how fast, is genuinely unresolved; production targets from companies like Tesla are treated with what one industry tracker bluntly calls "characteristic Tesla-timeline skepticism," meaning the gap between announced ambition and delivered scale in this sector has a documented history of running wide.

A newer wrinkle is the emergence of peer-to-peer robot rental platforms, built on the premise that as humanoid fleets scale, some owners will have more capacity than they can use — creating a secondary market where idle robots generate income for their owners while other operators rent capability without a six-figure capital commitment. This model is still emergent and unproven at scale; it's a reasonable economic hypothesis given the described asset dynamics, not a documented, functioning market yet.

Worth examining honestly

PROMETHEUS would tell you the BMW pilot and similar deployments are exactly what responsible automation looks like: narrow, task-specific, supplementing rather than replacing a human workforce while the technology and the labor market both adjust. GHOST PROTOCOL would add that "supplements rather than replaces, for now" is precisely the framing every automation wave uses in its early years, and the honest question worth tracking isn't this year's narrow pilot — it's the trajectory implied by $6 billion in committed capital chasing general-purpose humanoid labor, and where that trajectory points once the hardware and reasoning capability both mature past today's limits.

The Door This Opens

This door isn't asking whether robots will replace human workers tomorrow — the evidence says they won't, not yet, not broadly. It's asking whether the current "supplement, don't replace" phase is a stable equilibrium or a transitional one, and whether the workforce, the regulators, and the companies deploying this capital are preparing for the second possibility while living inside the first.