Nov 5, 2025 · 1h 36m · wtf
Humanoids Cost as Much as an SUV Now | Nikhil Kamath x Brett Adcock | WTF Online Ep 2 · Nikhil Kamath
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
Nikhil Kamath interviews Figure AI founder Brett Adcock to discuss the engineering breakthroughs, domestic safety challenges, and macroeconomic implications of general-purpose humanoid robots.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. Nikhil holds 15.2% of the talking time here. How this is scored →
Nikhil as informed peer 5.3 Guest teaching 5.8 Guest disagreement 2.6 Nikhil pushing back 2.8
speaking balance: gold is Nikhil, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 1:04:24 Dismissing Chinese robotics dominance claims
Brett directly rejects the widespread narrative that China leads robotics, arguing their hardware lacks autonomous AI capability and that US companies are far ahead.
Hardest push from Nikhil ▶ 1:01:35 Challenging the need for humanoids in structured factoriesNikhil pushes back on Figure's enterprise strategy by arguing factories can easily be redesigned for non-humanoid automation rather than expensive bipedal robots.
Biggest teaching moment ▶ 53:38 Educating on electric actuator torque density over hydraulicsBrett corrects the traditional engineering belief held by robotics pioneers that hydraulics are required for power, explaining recent motor torque density breakthroughs.
Nikhil holds their own ▶ 1:10:38 Macroeconomic framework on global services GDP versus AI ROINikhil demonstrates strong macro investor domain expertise by dissecting the $100 trillion global output, $50 trillion services economy, and the lagging revenue yield of AI investments.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | Nikhil as informed peer | Guest teaching | Guest disagreement | Nikhil pushing back | Why |
|---|---|---|---|---|---|---|
| Brett Adcock's Founder Journey: From Farming to eVTOL and Figure AI | 3 | 4 | 1 | 1 | Nikhil opens with friendly introductory questions about Figure AI and Brett's background. Brett outlines his path from Midwest farming to software exits, Archer Aviation, and humanoid robotics. | |
| The Pragmatics of Electric Aviation and Urban Air Mobility | 6 | 5 | 2 | 3 | Nikhil cites his investment in an Indian electric aviation startup and specific range specs to probe feasibility. Brett educates on energy density differences between kerosene and batteries, outlining the certification roadmap. | |
| The Humanoid Form Factor as the Ultimate Meta Problem | 4 | 6 | 2 | 2 | Brett explains the combinatorial complexity of 40 motorized joints yielding more potential states than atoms in the universe. Nikhil asks whether having humanoids roaming around feels dystopian or scary. | |
| Deconstructing Humanoid Hardware and Onboard Compute Architecture | 5 | 7 | 2 | 3 | Nikhil raises Moore's law limitations and questions onboard versus off-board compute. Brett details why 200 Hz closed-loop motor control demands edge compute and explains model quantization. | |
| Biological Energy Efficiency, Long-Horizon Tasks, and AI Voice EQ | 5 | 5 | 2 | 2 | Nikhil introduces an energy-input sandwich analogy to examine biological efficiency versus robotics. Brett details the complexity of long-horizon tasks like laundry planning and how native voice models can infer EQ. | |
| Voice as the Primary Interface for Artificial Superintelligence | 5 | 6 | 3 | 4 | Nikhil challenges voice as the ultimate UI by pointing out you cannot speak during meetings. Brett responds by arguing that current voice models are merely grade B and asserts all pre-AI computing hardware must be rebuilt. | |
| Hardware Differentiation, Edge Computing, and Robotic Fleet Learning | 6 | 6 | 2 | 3 | Nikhil shares his investment thesis in Nothing to discuss hardware commoditization versus differentiation. Brett breaks down fleet learning, explaining how logistics data unexpectedly transfers to domestic chores. | |
| Physical World Data, Domestic Safety, and the Westworld Aesthetic | 5 | 5 | 3 | 3 | Nikhil asks if Brett would leave his young kids with the robot and questions the humanoid aesthetic. Brett explains real-world safety limits and rejects childish toy aesthetics like googly eyes in favor of high-tech industrial design. | |
| Silicon Valley Narratives, Vision Servoing, and Electric Actuators | 6 | 7 | 3 | 4 | Nikhil references Yann LeCun and Boston Dynamics' historical defense of hydraulics. Brett details why recent torque density advancements in electric actuators rendered hydraulic systems obsolete. | |
| Industry Landscape: Debunking Chinese Dominance and Project Go Big | 6 | 8 | 6 | 4 | Nikhil probes Tesla's humanoid valuation, factory remodeling alternatives, and widespread narratives about Chinese robotics dominance. Brett forcefully rejects the China narrative, labeling tele-operated competitor demos as deceptive. | |
| Strategic Independence: Why Figure AI Ended the OpenAI Partnership | 5 | 6 | 4 | 2 | Nikhil asks why Figure AI terminated its high-profile partnership with OpenAI. Brett explains Figure possessed superior internal talent for embedded robotics models and wanted to avoid the perception of outsourcing AI. | |
| The Macroeconomics of Synthetic Labor and Frontier Tech Bets | 7 | 5 | 2 | 3 | Nikhil frames AI monetization using macro figures ($100T world GDP with $50T services) and pushes Brett to pick one non-Figure stock bet. Brett picks Waymo, and Nikhil validates the choice with a firsthand anecdote about a Waymo yielding to a firetruck. | |
| Entrepreneurial Playbook: Tackling Big Problems and Supply Chain Gaps | 6 | 6 | 2 | 2 | Nikhil asks what Indian entrepreneurs should build given lower labor costs and less risk capital. Brett advises pursuing massive binary outcomes and describes severe supply chain bottlenecks in custom motors and sensors. | |
| Advising Nikhil on Building an AI-Native Social Network | 5 | 6 | 2 | 2 | Nikhil seeks advice on building an AI-native social network out of India. Brett outlines his hiring philosophy (caring over pedigree) and suggests rapid prototyping with voice models within a week. | |
| The Future of Human Purpose, Executive Realities, and Farewell | 6 | 5 | 3 | 4 | Nikhil asks philosophical questions about whether human purpose was an illusion and whether Brett's drive stems from ego. Brett admits he hates losing and reflects candidly on grueling executive realities before closing the episode. |
Statements from this episode (35)
Adcock: Figure AI Aims to Replicate All Physical Human Capabilities
“And the goal here, it figures to be able to do everything a human can in the physical world.”
Adcock: Archer Aviation Flies Full-Scale Electric Aircraft Weekly
“I mean, we fly basically every week our full-scale aircraft at this point, fully electric, similar concept operations we'd want to do in market.”
Adcock: Paying Passengers Will Fly on Electric Aircraft Within 5 Years
“So I think within the next, like, five years, we'll be taking passengers, like, paying passengers around in, in air spaces here in the world on electric aircraft like this.”
Adcock: All Aviation Will Transition to Electric in Our Lifetime
“Like, my view is, like, all aircraft, like, all, all aviation, like, all air transport will move electric here. And I think hopefully in our lifetime”
Adcock: Billions of Humanoid Robots Will Exist Within Our Lifetime
“I think we'll, we'll see billions of humanoids on the planet in our, in, like, our lifetime, and they will do everything for us from things at home, like, doing the laundry, washing dishes unpacking groceries cleaning up the house to, we will see billions in t…”
Adcock: Figure AI designs its entire hardware and software stack in-house
“At Figure here, we design basically the entire hardware and software stack on the robot, almost entirely ourselves at this point. So that goes down to, like, the motors and battery systems, all the electronics, wiring, kinematics, and structures. And sensors.”
Adcock: Figure AI has shipped three robot generations in 3.5 years
“We're about three and a half years old now. We've shipped now three generations of humanoid robots. And we've been able to demonstrate that humanoids can do, like, useful human-like work now, fully autonomously, and in most cases, like, with neural nets all th…”
Adcock: True general-purpose humanoid robots are under 10 years away
“I don't, it's not something we'll solve in a month, but it's certainly not something that's 10 years away at this point.”
Adcock: Humanoids Must Run Real-Time Motor Neural Nets On-Board
“We need to run
A decent amount of the neural nets at, ah, kind of like, basically a couple hundred hertz, like a couple hundred times a second.
There's no way you're gonna be able to do that fully off board.
The robot, ah, would just be too slow.”
Adcock: Figure humanoids have run commercially at BMW for five months
“We have a commercial customer here in the U S we've now been running with for over five months now every day.
we put this out publicly.
It's with BMW.
That's a fully onboard onboard system that's doing all the reasoning onboard of what the robot should do.”
Adcock: Humanoid robots are far less energy-efficient than biological humans
“Today we're like far, like less efficient than a human. Humans are like the, are, we are like extremely efficient. Like, our, you know, biological neural nets are really efficient. The way, the, how much energy, or how much power we use to exert, like, walking…”
Adcock: Figure AI robots can work full human shifts today
“I wouldn't say it's a blocker today. Like, we can, we, like, we can do work In a lot of cases, almost the whole day, depending on what the task looks like and if it's not that, if it's, you know, a task that has a lot of locomotion and movements, we can do kin…”
Adcock: Figure 3 robot features three OLED screens and new audio system
“We have, like, in figure three, we have three screens. We have a front screen, like, an OLED. We have, like, two side screens. We can, like, I'll talk about that. We have a speaker and mic, a new speaker and mic system that's, like, really, actually really goo…”
Adcock: Voice AI will approach human Turing tests within 12–18 months
“We will see in 12 to 18 months voice models approaching human Turing tests, and that's not just from an IQ perspective and information retrieval, it's EQ as well, as we just talked about in naturalness.”
Adcock: AI Hardware and Humanoids Will Replace Phones and Computers
“Yes. Every phone and computer was built pre-AI. The, All the hardware we have today in the world needs to be completely rebuilt from scratch.”
Adcock: Humanoids will combine off-board cloud brains with fast on-board models
“I think you will certainly have off board reasoning. I think you will have, I think you'll have both. I think you'll have off-board, like, a big brain off-board that you need to talk to that can do very complex maybe planning and reasoning. You'll have on-boar…”
Adcock: Deep Learning Transfers From Logistics Tasks to Folding Laundry
“We are taking data from all the robots, and we are consulting that out to, like, our servers here at Figure, and we're training on that. And what we're seeing now is we're seeing, like, incredible transfer from other robot data that's helping the current robot…”
Adcock: Humanoid robotics fleet scaling instantly propagates skills from one robot to millions
“Once, like, one robot knows how to do, like, logistics, That's the same neural networks we can put to every single robot in the fleet, and they can all do it. And so you have this weird scaling law here for humanoids where once, like, one robot knows how to do…”
Adcock: AI has exhausted internet text tokens, requiring physical world data
“We've kind of, like we've, like, tapped out all the, kind of, tokens on the internet at this point. So we think, like, being able to explore the physical world is, like, extremely important for, like, overall intelligence.”
Adcock: I Won't Let Figure Robots Roam Unsupervised Around My Kids
“I've had a robot in my house now for like a few months. And You know, my kids have been there, but we, like, we monitor the robot. So we're not, like, we're not, like, we're, like, we have somebody with the robot. Like, we're not letting the kids go up and hug…”
Adcock: Figure robots do not have facial expressions or human-like features
“Our view right now is that we are trying to go out and do useful work. We're not trying to fool another human that we're a human from, like, a entertainment perspective or maybe companion perspective. So we've not designed the robot with, like, facial expressi…”
Adcock: Actuator and battery improvements allow fully electric humanoid robots
“We've now had torque density and power density of actuators and battery systems get good enough where you can run, like, fully electric systems instead of hydraulic systems, and that makes the robot smaller, lighter weight, quieter, faster like, you can run, t…”
Adcock: Most humanoid competitors deceptively rely on human teleoperation in demos
“And what you're seeing now in the world is like, I would say most major competitor to us at this point are putting out most of their content and updates teleopt from a human. And I think it's like perhaps some of the most deceiving things I've ever seen.”
Adcock: Figure AI and the US lead all Chinese humanoid competitors
“The U.S., and particularly Figure, is ahead of every single Chinese company I've seen globally, of what I know publicly, in terms of capabilities. Like, watch the robot do advanced things with neural nets autonomously over long horizons. And the answer is, lik…”
Adcock: Humanoid robot manufacturing resembles consumer electronics, not cars
“Automotive manufacturing is like the most complex systems I've ever seen in my entire life all put together to manufacture cars. It's like hundreds of times harder than manufacturing consumer electronics. Like, we manufacture consumer electronics a lot by hand…”
Adcock: Figure AI ended OpenAI partnership after a year to build in-house
“OpenAI led my series B. We generally were working on some kind of advanced models with them. Or maybe a year or two, a year or so, about a year. And I figure ultimately chose to kind of, ah, leave them and go in our own direction.”
Adcock: OpenAI partnership hurt hiring by creating perception of outsourced AI
“It also became hard, like, hiring people when we had a partnership like that, where folks felt like we were outsourcing our AI, which we weren't doing, and we were actually doing all of it mostly ourselves internally.”
Adcock: Figure AI evaluates 5 to 20 new models daily on hardware
“I sit here and watch us do generally anywhere between, like, five and 20 evaluations of new models every single day. And, like, we've been doing that now for a year and a half, two years.”
Adcock: Humanoid robotics will create the world's largest company by far
“My view is that there's going to be a generational company to be built that We'll be able to put, ultimately, billions of robots on the planet that will build, like, the largest company we've ever seen in the world by a long shot.”
Adcock: Waymo is super underrated at a $50 billion valuation
“I would probably pick Waymo as my pick. I think I, you know, I think it's like a fifty billion market cap super underrated.”
Adcock: Figure AI has over $1 billion in cash
“Like, for us, like, we have, like, we have, like, over a billion dollars of cash.”
Adcock: Hiring experienced big-company executives is a flawed playbook
“I've always I think my, like, the conventional wisdom of everything is to go out and hire somebody, like, really experienced that makes you feel really good from a really good background that's at a successful company. And I found, like, that playbook is just …”
Adcock: Meta's AI lab approach is going to collapse
“It's, like, the opposite of what Meta, I think Meta's doing that right now, right? Look at the Meta, like, a super intelligence lab. It's, like, you're putting, like, I mean, it's just gonna, it's just immediately gonna, it's gonna immediately collapse. It's n…”
Adcock: Goods and services prices will collapse toward zero with synthetic agents
“The prices of goods and services are, will slowly start collapsing, basically, anatomically approaching zero.”
Adcock: Robot costs will drop dramatically over the next five years
“You'll have this dramatic reduction here over the next, like, five years as we approach higher volume manufacturing, and then you'll have another step down once robots can fully build robots in the loop together, and you remove all human labor from the whole m…”