Doug Field Leaves Ford for a $3.4B Rivian Spinoff — A Labor-Market Tell for the Embodied-AI Trade

On Oct. 13, 2026, onstage at TechCrunch Disrupt, Rivian founder and CEO RJ Scaringe announced that Mind Robotics — the industrial-AI and robotics company he spun out of Rivian ten months ago — had hired Doug Field as president. Field will run day-to-day product, engineering and operations, leaving Scaringe in a chair role. The hire is the most senior OEM-production executive yet to cross over to a private physical-AI startup, and it lands a deliberate six months after Field quietly left Ford as the auto maker dissolved the unit he led.

What Mind Robotics actually bought is a specific blend of skills that almost no other human in the U.S. labor market carries: software-defined-vehicle architecture, automotive-grade quality systems, and a real factory ramp. Field was senior vice president of engineering at Tesla during the Model 3 production hell; ran Apple's special-projects group after the company shut down Project Titan in 2024; and then oversaw Ford's electric-vehicle, digital and design function across the F-150 Lightning, Mustang Mach-E and a software stack that never quite shipped. That stack is now exactly what industrial AI needs to graduate from demos to line work.

Mind Robotics was created in November 2025 with a $115M seed round led by Eclipse, with Rivian also investing. A $500M Series A followed in March 2026, co-led by Accel and Andreessen Horowitz, valuing the company at roughly $2B. In May 2026, Kleiner Perkins led a $400M round that pushed the valuation to $3.4B, per Reuters. Across those rounds, the same institutional pattern shows up: top-tier Silicon Valley venture funds plus the venture arms of Volkswagen and Salesforce — i.e., a U.S. car buyer, a German OEM that already has a $5.8B JV with Rivian, and a software giant that runs call-center automation at scale.

The parent is executing. Rivian reported Q3 2026 deliveries of 19,248 vehicles on Oct. 2, a46% jump year-over-year and roughly 1,300 units above Wall Street estimates, per CNBC. Q2 2026 revenue of $1.658B grew 27% year-over-year on the first external R2 shipments and produced a positive gross profit of $179M, per the company's Jul. 30 earnings release. The R2 program officially started production at the Normal, Illinois plant on April 22, 2026, with a designed capacity of up to 155,000 units per year. None of that has translated into a working multiple: Rivian trades around $13.40 with a market cap near $16.2B — barely half of what the company's Q2 enterprise value implied a quarter ago — and a CFO transition announced Aug. 27 sent the shares down 6.3%.

The CFO transition announced Aug. 27 — Claire McDonough stepping down Oct. 30 after nearly six years, with Derek Mulvey named successor — is unrelated to Mind Robotics on paper but identical in pattern: another senior Rivian operator is leaving the EV business, and the market has no clean way to mark the residual value of the 35–40% stake the company is widely estimated to retain in Mind.

Mind's $1B raised in 10 months is not an outlier — it is the floor for serious embodied-AI capital in 2026. Figure raised more than $1B at a $39B valuation in September 2025; Apptronik closed a $520M Series A extension in February 2026 at roughly a $5B valuation; Physical Intelligence raised $400M in late 2024 at a $2.4B valuation; and 1X Technologies is publicly targeting a $10B valuation. The capital is no longer flowing toward humanoid demos alone — it's flowing toward companies with at least one anchor industrial customer. Mind's anchor is Rivian; Apptronik's is Google's DeepMind-aligned logistics pilots; Physical Intelligence's is a BMW partnership; Figure's is BMW manufacturing deployments.

The Field hire crystallizes a rotation that has been quietly underway for two years. Capital is leaving auto incumbents and landing on three layers of the physical-AI stack: the compute layer (NVIDIA), the cobot and mobile-robot layer (Teradyne via Universal Robots, and Yaskawa/FANUC on the industrial side), and the warehouse-automation layer (Symbotic, AutoStore, Rockwell). On the other side, the legacy OEM that just lost Field is shrinking its own robotics footprint: ABB agreed in October 2025 to sell its entire robotics division to SoftBank for $5.375B, with the deal expected to close in mid-to-late 2026.

Two distinct sell-side frames now compete. Bulls argue that NVIDIA's Jetson and Omniverse stack becomes the default substrate for any industrial foundation model — including Mind's — which means each $1B in private robot capital eventually pulls a margin-rich order through NVIDIA's data-center and edge-AI lines. Bears point out that ABB's exit proves even the largest installed-base industrial-robot company could not monetize the physical-AI transition in-house, and that Symbotic's 500x trailing P/E and Teradyne's 53x are pricing in much of the same optionality the private market is paying for in startups like Mind.

In the days and quarters ahead, the catalysts are execution, not narrative. Mind's first deployment — Dexterous manufacturing cells inside the Rivian Normal, Illinois R2 line — has to ship with the unit's published capacity ramp from April 2026 toward 155,000 vehicles per year, or the foundation-model thesis loses its single most important training-data source. The next public milestone is Rivian's Q3 2026 earnings call on Oct. 29, where investors will likely press for any incremental Mind revenue contribution or updated stake disclosure. On the public side, NVIDIA's GTC fall event and Teradyne's Q3 print are the closest windows into whether industrial-AI capex is turning into orders.

Over the one-to-three-year horizon, the structural call is sharper. If Rivian succeeds in converting its R2 ramp and the Volkswagen JV into a stable foundry-style production base, Mind becomes a credible second-act option for any OEM executive watching the same talent flow; if not, Mind pivots into third-party deployments and the $3.4B valuation becomes the new benchmark rather than a special case. Either path keeps the demand-side case for NVIDIA, Teradyne and Symbotic intact, and either path continues to drain engineering talent from the Ford and Tesla benches — which is the original tell.