My Small Satellite Wager on Tomorrow's Tireless Workforce

A billion robots sounds like a movie poster. Yet MorganStanley penciled it into the calendar for 2050: nearly a billion human-like robots and a $5T market around them. When I hear numbers that large, I check my wallet first and ask which single ticker lets me sit at that table. My answer this time is one door: the BOT fund, a closed-end vehicle trading on the Nasdaq exchange that owns stakes in private robotics firms. In this piece I take that door apart, bolt by bolt, and ask whether behind it sits a price or a prize.

Let me drag the phrase physical AI down into my kitchen. A chatbot can hand me a sandwich recipe; spreading mayo from the jar with a knife in my hand is another craft entirely, how hard to press and which angle to turn my wrist, all decided in milliseconds. That is what physical AI means: a robot that holds the bottle without dropping it and pours without spilling. It sounds simple because our bodies do it without thinking; a robot has to learn physics from scratch.

Why is this speeding up now? Three years ago chatbots changed desk jobs; office crews watched their daily load shrink. As our guest put it, the confidence boom in knowledge work is now walking into the workshop. Arms that flip parcels in depots and fit parts at benches are living through the same jump. Words got solved first; hands are next.

Let me run the math backwards. The planetary pool of physical labor is worth about $50T. A robot that skips sleep and sick days, runs three shifts, and covers the output of three to four people looks like free overtime to an employer. At giants like Walmart or General Mills, half to three-quarters of revenue goes to human muscle; a firm that trims that bill can double or triple its profit. Whenever I pick a stock, I ask one question: will this business grow by spending less on muscle?

I cross-checked the figures from two desks. MorganStanley expects a $5T market by 2050 with close to a billion robots, 90 percent of them in factories and commercial sites. China leads with 302 million units while America sits near 78 million, and the real acceleration lands in the late 2030s. CNBC frames $4.7T of sales as twice the combined revenue of today's 20 biggest carmakers. Goldman sketches $38B for 2035, BofA ten million units; everyone sees the same ocean, they only disagree on the tide table.

Robots will not drop through our chimneys; the order is fixed. First come repeat jobs in depots and at benches, Figure.ai units are already rehearsing exactly that inside logistics hubs. Then come thinking jobs in diners and hospitals; the home waits last. The reason is plain arithmetic: a factory happily pays $100-150K a year for a robot, runs it around the clock, and earns the money back. Who pays the same sum to have floors swept at home? Only the richest. So the calendar reads factory first, home later.

Now the star of the basket: Figure.ai. Founder Brett Adcock previously built two companies and took one public, arriving from the flying-car business, a record of inventing devices from zero. He staffed up from the kitchens of Boston Dynamics, Tesla, and DeepMind, gathering the world's hundred-odd masters of each narrow craft, hands, actuators, fleet control, under one roof. For me, the crew comes before the balance sheet.

The numbers move fast too: Figure.ai passed $1B of funding at a $39B valuation in September 2025, led by Parkway and Brookfield, with large processor and software houses joining, and the cash feeds the Helix intelligence setup plus the BotQ plant. The BMW trial, though, taps the brakes: Fortune reports a single robot carried parts on night shifts for a long stretch, and the fleet and end-to-end-operation talk looked overstated. I take that warning seriously, a rehearsal is not a deployment.

So why are such bright companies not listed? Apple, Amazon, and Nvidia once floated near $1B valuations; today OpenAI, Anthropic, and SpaceX stay private toward the $1T line. The causes are clear: reporting, boards, and compliance weigh heavier on listed firms, while VC pools, buyout money, corporate treasuries, even sovereign funds flood the private side. The average wait for a listing stretched from 8 years to 14. The train is leaving, but the platform is private, and vehicles like BOT sell tickets to it.

The RoboStrategy basket splits three ways. The first pocket holds general-purpose humanoids like Figure.ai. The second is my favorite: arms and helper cobots that have crawled blindly from point A to point B in factories for years. They were blind until now; once they gain systems that read their surroundings, the same steel grows smart. Most jobs need no legs anyway, a wheeled base does the trip. So the second pocket carries a stock story as large as the first.

The third pocket is the supply chain: actuators. An actuator is the muscle and joint of a robot, built from motors, gears, and power electronics. Dozens sit in each unit and make up 30-50 percent of the hardware bill. Output today counts in hundreds of thousands a year; tomorrow it must count in the billions, just as processors sold out first in the AI wave and memory parts next. I love such bottleneck stocks; bought with price discipline, they can carry a portfolio.

A fourth pocket holds applied plays: Path Robotics does AI welding. America lacks hundreds of thousands of welders a year; nobody wants the mask. HII, the shipbuilding giant, signed a multi-year production deal in August 2026 with Path and GrayMatter worth up to $900M. Blocks of ships welded by robots, in other words. That deal stamps the moment the robotics wave turned from talk into work, and I collect stamps.

Now let me open the vehicle itself. BOT is a closed-end fund: it trades like a stock on the exchange, yet unlike an ETF it creates and redeems nothing day to day. It sits on stakes in unlisted companies, the manager barely trades, and holds for years. The logic copies the early Facebook backers' lesson: when you catch a rare giant, the only job is to hold. For an impatient investor like me, that frame is medicine, no panic taps on the screen.

Two engines grow the per-share value. The first is the basket itself: if names like Figure.ai grow, the fund's stake value climbs. The second is the premium game: when fund shares trade above their inner value, fresh shares get sold, cash lands in the vault, and the cash buys more robotics stakes. Picture one Figure.ai share per fund share compounding into three through such rounds. TransDigm and Constellation ran the same play for years, buy cheap in private deals, revalue at public multiples. I admire that second engine, though I doubt the premium lasts forever.

Let me lay the risks on the table too. First, plant capacity: a million robots do not snap out of fingers; land, halls, craftsmen, and supply lines take years. Figure.ai, Tesla Optimus, and Apptronik all expand halls, yet calendars can slip. Second, the intelligence schedule: if the jump from today's level to the next takes two or three years instead of one, valuations will sweat. Third, market mood: should the premium flip into a discount, the share price decouples fast from inner value. I file all three under short-term bumps, not a broken thesis, but I keep the position small.

Then there is the NAV lag. Private stakes get no daily print; they sit marked at the last funding round, sometimes with a small haircut. Rounds stand 12-24 months apart, like guessing today's price of a house sold three years ago. So the fund's market quote and its inner value diverge; the Nasdaq screen mirrors the impatient, the NAV mirrors the patient. Holders learned the per-share figure for September 2026 was $11.35. The fee side reads honest: a 2.5 percent yearly management charge looks modest beside the classic VC two-and-twenty. One-off listing costs will melt away and the ratio should sink toward the management fee.

Final question: what does the world look like in five years? My read: hundreds of thousands of robots in factories within two to three years, then millions; tens of millions of units and first guests at home in the early 2030s. A single million robots alone can justify $1T giants. What do I do? I made the BOT fund a small satellite in my portfolio, adding on discounts, waiting when the premium foams. This is no sprint, it is a marathon, and the water stations are marked: balance sheet, premium, capacity.