explainx / blog
Figure AI CEO Brett Adcock confirmed robots now outnumber humans at the company. Here's what the headcount chart, 1/hour production rate, and BMW deployment numbers actually mean.
Jun 20, 2026
AI is automating jobs, robots outnumber humans at Figure, and 97,000 tech workers were laid off last month. Here's the practical playbook for navigating what comes next.
Jul 10, 2026
1X calls NEO's new hands "an API to the physical world" — 25 force-controlled DoF, quasi-direct-drive tendons, tactile shear sensing, and a 10K/year production line. explainx.ai breaks down the July 9 launch with hosted demo clips.
Jun 29, 2026
From garment factories in Tamil Nadu to homes in Hyderabad, thousands of Indian workers are strapping on head-mounted cameras to record their daily tasks. The footage trains the Large Behaviour Models powering the next generation of humanoid robots — robots that may ultimately replace the same workers doing the recording.
On June 20, 2026, Figure AI CEO Brett Adcock posted four words on X that mark a genuine first: "robots now outnumber humans at Figure."
The chart he shared makes the trajectory viscerally clear. From Q2 2022 through most of 2024, the human line climbs steadily and the robot line barely registers. Then, sometime in early 2025, the robot curve starts accelerating. By Q2 2026, it crosses 750—well past the human employee count, which has stayed relatively flat around 200–250.
Update — July 10, 2026: 1X NEO shipped 25-DoF force-transparent hands with a 10K hands/year production target — home dexterity hardware vs Figure's factory BotQ ramp.
It is the first time a company of meaningful scale has publicly confirmed its deployed robot fleet exceeds its human headcount. That Figure AI is the company where this happened first is not coincidental—it is, after all, the company building the robots. But the self-referential nature of the milestone is exactly what makes it a pointed early data point for where this is heading.
The headcount vs. robots graph Adcock shared runs from Q2 2022 through Q2 2026. A few things stand out:
The human line is flat: Figure has maintained a relatively lean headcount—approximately 180–250 employees—consistent with a hardware-AI company optimizing for capital efficiency over headcount growth. You do not scale humanoid robot production by hiring more people; you scale it by building better factories.
The robot line is exponential: Near-zero through 2023 and most of 2024, then a steep climb through 2025 that turns nearly vertical in Q1–Q2 2026. This tracks directly with the production ramp at Figure's BotQ factory in California.
The crossover happens in Q2 2026: The lines intersect somewhere in the first half of 2026—and by the time Adcock posted, the robot count had already pulled significantly ahead.
The chart is not showing deployed units at customer sites exclusively—it likely reflects total Figure robots in existence (factory, customer deployment, internal testing). But the direction is unambiguous regardless of how you slice it.
The robot surge in the chart is directly explained by what happened at Figure's BotQ factory between late 2025 and mid-2026.
October 2025: Figure unveils the Figure 03—its first robot designed from the ground up for mass manufacturing rather than demonstration. At 5'8", 61 kg, with 48+ degrees of freedom, a 20 kg payload capacity, and wireless foot-charging, Figure 03 is built to be made at scale, not just to work at scale.
February 2026: BotQ is producing approximately 60 robots per month.
April 2026: Output has climbed to 240 per month—a 4x increase in two months.
May 2026: Figure publicly announces it has ramped from one robot per day to one robot per hour in under 120 days—a 24x throughput improvement. The BotQ facility is designed for 12,000 units per year in its initial configuration, with a stated path to 100,000 per year.
June 20, 2026: Robots outnumber humans at the company.
The production math is consistent with the chart. At one per hour, Figure produces roughly 720 robots per month. Even accounting for ramp-up time and the earlier slower rates, reaching 750+ total robots by Q2 2026 fits the publicly disclosed production trajectory.
The production numbers matter less than the deployment numbers—robots that are working, not sitting in a warehouse.
BMW Spartanburg (Figure 02, 2024–2025)
Figure 02 completed an 11-month commercial trial at BMW's Spartanburg, South Carolina plant. The results, by any measure, are significant:
These are not demo numbers. BMW did not let a humanoid robot anywhere near a production line without extensive validation—and kept it there for nearly a year.
BMW Spartanburg (Figure 03, 2026)
Following the Figure 02 trial, BMW has deployed 40 Figure 03 units at Spartanburg, with phased expansion planned across additional workstations through 2026 and 2027. Pilot programmes at BMW's German facilities—Munich, Regensburg, and Leipzig—are also in progress.
Hyundai and Google DeepMind
Both companies have reserved 2026 production capacity from Figure's BotQ factory. The nature of the Google DeepMind partnership is particularly interesting—it suggests the robots are being used for research into real-world AI behavior and physical-world data collection, not just industrial automation.
The hardware milestone would be meaningless without the AI to run it. Figure's Helix-02 collaboration demonstration in May 2026 showed what the underlying AI can actually do: two robots, running a single shared Vision-Language-Action (VLA) policy, coordinating to clean a bedroom—making a bed, hanging clothes, navigating around each other—without central planners, without explicit message passing, inferring intent from motion alone.
That is the AI stack running in every Figure 03. Helix processes raw camera pixels and proprioception, outputs torques directly, and updates in real time—no pre-built object models, no hand-coded manipulation logic. The same policy that learns to tidy a bedroom generalizes to unfamiliar industrial environments with minimal re-training.
The VLA approach is what makes the 1-per-hour production rate meaningful: you are not building a bespoke robot for each deployment. You are building one platform and training one policy that adapts.
The reactions to Adcock's post split roughly into three camps:
"This is amazing progress" — The people pointing out that Figure built its way to this milestone in under four years, from founding (2022) to robot-majority workforce (2026), and that the production ramp is genuinely unprecedented in the humanoid space.
"Is that robots at work or robots built?" — A legitimate question. The chart does not specify whether the robot count reflects deployed units only or all Figure 03 units in existence. Given the production rate and BMW deployment numbers, the answer is probably some of both.
"This is a turning point" — The more philosophically charged reactions, including the widely-quoted X post: "Good luck humanity because this is a turning point." This is probably both true and overstated simultaneously.
Here is what the milestone concretely means:
For Figure: The production ramp is working. The bottleneck is no longer "can we build robots" but "can we find deployment opportunities fast enough." With BMW, Hyundai, and Google DeepMind on the books, that constraint is also relaxing.
For the industry: Figure is the public proof of concept that humanoid robot production can scale like other manufactured goods. If Figure can do 1/hour now, Tesla Optimus and Agility Robotics will be under pressure to match that pace.
For the broader economy: The first company where robots outnumber humans is, pointedly, a robot company. This will not stay true of robot companies exclusively. The question is how fast the crossover extends to other industries—logistics, food service, elder care—and on what timeline.
The industrial deployments are the proof of commercial viability. But Brett Adcock has been consistent about where Figure is ultimately headed: the home.
"The home is coming. The home is like single-digit years away from being a place where humanoid robots can do useful work."
This is the scenario that makes the June 20 milestone look small in retrospect. BMW Spartanburg employs structured, predictable environments with well-understood tasks. The home is the opposite: cluttered, unstructured, socially complex, and different in every household. Getting a robot to sort laundry on a factory floor and getting it to sort laundry in your apartment are meaningfully different problems.
The Helix-02 bedroom demonstration was specifically staged in a home environment—a deliberate signal about where Figure's research attention is pointed even while its revenue comes from industrial contracts.
The VLA policy approach is well-suited to this transition. Unlike rule-based systems that need reprogramming for every new environment, a model that learns from pixels generalizes. The question is whether it generalizes enough, fast enough, cheaply enough to make home deployment commercially viable in the timeframe Adcock is describing.
Figure's Series C raised $1 billion at a $39.5 billion valuation—putting it in the same tier as SpaceX and OpenAI among private companies. That valuation is a bet on the home scenario, not just the BMW contract.
The industrial revenue is real but limited: even at full Spartanburg deployment, 40 Figure 03 units at likely $100K–$200K per unit per year is tens of millions of revenue. The $39.5B valuation implies investors believe Figure captures a meaningful share of a market that does not yet exist at scale.
What justifies that belief:
The risk: everything above assumes the home transition happens in the "single-digit years" Adcock cites. If it takes fifteen years, the valuation math changes considerably.
The honest answer is: it depends on which workers, in which industries, on what timeline.
The near-term pattern from BMW is substitution of specific physically demanding tasks—parts handling, repetitive assembly movements—not wholesale job elimination. Figure robots at Spartanburg work alongside human workers, not instead of them. The productivity gain comes from robots doing the physically taxing portions of jobs that are hard to fill and lead to high injury rates.
The medium-term pattern, if Figure scales to home environments, is more disruptive. Domestic service work—cleaning, elder care, childcare—employs tens of millions of people globally and is structurally difficult to automate with traditional robotics. A capable general-purpose home robot changes that calculus.
The economic implications of AI-driven workforce changes are already being debated actively across industries. The Figure milestone adds a physical-world data point to what has been primarily a software-and-white-collar conversation.
For workers specifically: the industries with the highest Figure deployment risk in the next three to five years are structured industrial settings with repetitive physical tasks. The industries with the highest risk in the five to ten year window, if the home transition happens on Adcock's timeline, are domestic and care services.
June 20, 2026 is a date worth noting. Not because robots are about to replace everyone—they are not, not this decade—but because the first company where the crossover happened is now a data point, not a prediction.
Update (July 9, 2026): Mistral's Robostral Navigate tackles a different slice of the robotics stack—map-less language navigation from a single RGB camera (8B model, 76.6% R2R-CE unseen)—complementing humanoid VLA systems like Figure's Helix rather than replacing them.
Figure AI built its way to robot-majority in four years, starting from a blank-sheet design in 2022. It did it with a lean human team, a purpose-built factory, and a bet on Vision-Language-Action AI that appears to be paying off. The BMW proof of concept held. Production is scaling. Customers with real industrial operations are paying real money.
The chart Brett Adcock posted on June 20 will look different in two years. The question is whether the robot line is at 5,000, or 50,000—and whether those robots are still mostly on factory floors, or whether some of them are in your home.
How many robots does Figure AI have compared to employees? As of June 20, 2026, Figure AI CEO Brett Adcock confirmed that the company's robot count has surpassed its human headcount for the first time. Figure has approximately 180–250 human employees. The headcount vs. robots chart Adcock shared shows the robot count accelerating steeply through Q1–Q2 2026, crossing the 750 mark—well above the human employee line.
What is Figure AI's current production rate? As of May 2026, Figure AI produces one Figure 03 humanoid robot per hour at its BotQ factory in California—up from one per day just 120 days earlier. That is a 24x throughput improvement. Monthly output grew from roughly 60 robots in February 2026 to 240 in April 2026.
Where are Figure robots currently deployed? Figure 02 ran an 11-month trial at BMW's Spartanburg plant, helping produce more than 30,000 BMW X3 vehicles. Figure 03 has since been deployed at the same facility with 40 units, with expansion planned across Spartanburg workstations through 2026–2027, plus pilot programmes at BMW's German plants. Hyundai and Google DeepMind have also reserved 2026 production capacity.
What is Figure AI's valuation in 2026? Figure AI raised $1 billion in its Series C round at a $39.5 billion valuation, with over $1.9 billion in total funding across all rounds.
What is the Figure 03 robot? Figure 03, unveiled in October 2025, stands 5'8" tall, weighs 61 kg, carries 20 kg payloads, and walks at 1.2 m/s. It has more than 48 degrees of freedom and runs on Figure's proprietary Helix AI platform. Wireless charging is built into the feet.
Should workers be worried about humanoid robots replacing jobs? The near-term pattern is task substitution within structured industrial roles, not wholesale job elimination. Figure robots currently work alongside human workers at BMW, handling physically demanding tasks. The longer-term trajectory—home environments, care services—is where the picture becomes genuinely uncertain. Brett Adcock has said the home is "single-digit years away" from meaningful robot deployment.
What does it mean that Figure AI itself is the first company where robots outnumber people? It is both a milestone and a signal. As a milestone: the first time a company of meaningful scale has publicly confirmed its robot fleet exceeds its human headcount. As a signal: the company building the robots is demonstrating that the economics of robot labor can scale faster than human hiring—at least in controlled production contexts.
