Tesla scrapped two cars to build a failing robot as rivals already ship functional units

Tesla’s Fremont plant is being re‑tooled to produce a “1 million‑unit‑per‑year” line of Optimus humanoids, a shift announced on Jan. 28, 2026 when Elon Musk told investors that none of the company’s robots were performing useful work in its factories. Musk’s admission directly contradicts earlier public statements: in June 2024 Tesla claimed two Optimus units were already operating autonomously on the assembly line, and by early 2025 he had promised “several thousand” functional bots by year‑end with a target of 10 000 units. The latest filing shows the company has instead halted production of its Model S and Model X—its two oldest vehicle platforms—to free up the same assembly lines for robot manufacturing, a move Musk described as an “honorable discharge” for the cars that built Tesla’s reputation (Moth, Feb 24).

Boston Dynamics, by contrast, unveiled a production‑ready Electric Atlas at CES 2026 on Jan. 5, and the unit is already allocated to paying customers. Hyundai secured the first fleet for its Robotics Metaplant Application Center, while Google DeepMind is integrating Gemini Robotics foundation models as Atlas’s brain, giving the robot multi‑step reasoning capabilities in unstructured environments. Atlas can lift 110 lb, reach 7.5 ft, and operate from –4 °F to 104 °F, with Hyundai planning factory‑floor deployment in 2028 and full‑assembly work by 2030 (Moth, Feb 24). All 2026 units are “already spoken for,” indicating a fully booked order book and no need for speculative production runs.

Figure AI has taken a different route, building a self‑replicating production line that lets its humanoids assemble subsequent robots. The BotQ facility, announced late 2025, can output 12 000 robots per year and aims for 100 000 within four years. Figure’s Helix AI system assigns humanoids as material handlers between stations, eliminating traditional conveyor belts and dramatically flattening the cost curve. The company raised $675 million from Nvidia, Jeff Bezos, OpenAI and Microsoft, and plans to scale to 20 000 robots in 2026 (Moth, Feb 24). This “robots building robots” strategy directly addresses the efficiency gap Tesla is currently facing, as its Optimus units have yet to demonstrate any material contribution on the shop floor.

Tesla’s pivot also coincides with a $2 billion investment in xAI, Musk’s separate artificial‑intelligence venture, which is currently being sued by Tesla shareholders for allegedly siphoning AI talent and resources. xAI has told its own investors that it will “develop self‑sufficient AI to power robots like Tesla’s Optimus.” Five days after the investment, SpaceX acquired xAI in a $1.25 trillion merger—the largest in history—yet the deal excluded Tesla, leaving the automaker without direct access to the AI assets it just funded (Moth, Feb 24). The juxtaposition of a massive AI infusion and the simultaneous shutdown of two flagship vehicle lines underscores the high‑risk gamble Tesla is making on a robot that, by its own CEO’s admission, does not yet perform any useful tasks.

Industry analysts note that while Tesla’s brand cachet remains strong, the company’s robot strategy now lags behind competitors that are already shipping functional units or have built self‑sustaining manufacturing ecosystems. Boston Dynamics’ Atlas is already in commercial use, and Figure AI’s self‑replicating line promises exponential scaling that could undercut traditional automation costs. Tesla’s reliance on a single factory conversion, combined with the legal and financial entanglements surrounding xAI, creates a precarious path forward. If Optimus cannot demonstrate material utility by the end of 2026, the “1 million‑unit” production target may remain a speculative headline rather than an operational reality.