If you were young enough to read in the 80s, the following statement by Bill Gates shouldn't surprise you.
Bill Gates invented one critical thing for the entire software industry, and that is:
You don't own the software we make. You license it.
When you purchase software, you generally buy a license to use the software rather than owning it. This means you have the right to use the software according to the terms set out in the license agreement.
Elon Musk likely nailed one thing for the hardware industry, and that is:
A humanoid, for every human on Earth
Followed by, if you have a humanoid and you don't have a digital brain to install in it, your humanoid is as good as my tuna can. Collect screws in a garage. #truestory
Some believe that Brett Adcock (37) is on a path to be the next Elon Musk (53). If you never heard of Brett likely is because he's the type of talent that you don't see coming until he is already in your living room, pitching the next big thing.
The man is driven by an insatiable passion for whatever he sets his mind on. Adcock has a background in business administration from the University of Florida and a history of entrepreneurial success.
Adcock's journey in the realm of technology and entrepreneurship began early. Raised on a family farm in Central Illinois, he ventured into building web companies as a teenager, displaying an early knack for innovation and business. He has since dedicated over two decades to building technology companies, starting in the software and internet space and later moving into AI and hardware.
One of Adcock's significant accomplishments is founding Vettery, an AI-based talent marketplace. Established when he was just 26, Vettery grew rapidly, expanding its network to thousands of companies and aiding numerous individuals in finding their ideal jobs. The company's success culminated in its acquisition by The Adecco Group in 2018 for $110 million.
Adcock made a notable mark in the aerospace sector by founding Archer Aviation in 2018. This venture focused on developing electric vertical takeoff and landing aircraft to address urban mobility and sustainable transportation challenges. Under his leadership, Archer Aviation evolved into a leading company in its field, achieving substantial milestones, including a public listing on the NYSE with a valuation of $2.7 billion and a significant commercial agreement with United Airlines.
In 2022, Adcock embarked on a new endeavor with the founding of Figure, an AI Robotics company focused on developing a general-purpose humanoid robot. This initiative aims to address the labor crisis in essential industries by introducing autonomous humanoid workers capable of operating alongside humans in various sectors.
Just like Mr Musk, he dumped all he had accrued from a previous succe$$ story into the next venture; when they say "put skin in the game," he is a leading example of that. A very supportive wife and two kids are the only major difference from Mr. Musk's that he tends to ingravidate whatever he touches. 😅 The man walks the talk when it comes to solving falling population numbers.
I know Brett exist because during my past startup lives I often needed to navigate the nightmare of jargon and "rules" in finances. Years ago, I found an incredibly valuable resource for me and my peers. Street of Walls. I wanted to send the author a basket of Italian sweets as a thank you for putting all that knowledge for free out in public, and then I found out that he was the CEO of a company going IPO. Ok, fine!
The man tried to get a job at JP Morgan didn't make the interview, so he decided to learn all about the space, in the process he wrote and posted onlineThe man tried to get a job at JP Morgan but didn't make the interview, so he decided to learn all about the space; in the process, he wrote and posted online, for free, everything that he learned on finances. The site is still incredibly popular these days.
By the way, he got the job, and very shortly, he said naaaaahh, I am good. Let's build, with no prior knowledge, hooks, and enough money for the job, an electric helicopter (eVTOL) and take the company to IPO. And, reportedly, he prepared most of the IPO paperwork by himself. How's that for: I am not wasting my knowledge, period. Attitude.
I wanted to proper document some of the most salient parts of Brett Adcock rising to the more popular eye because by all accounts Brett is on the verge of making a positive dent into humanity at the same scale that Musk did thus far with Space and electrification. Both will coexist in the market and public arena, but both companies' cultural latitudes are drastically different.
Humanoids will do useful work that humans today either are increasingly unwilling to do or to do for a long time - some tasks are too dangerous for humans.
Do not make any mistake, Figure is an AI company. It's in the incorporation name, it's in the nature of the work that makes the robot accomplishing useful tasks. The fact that they make a machine that moves is a side effect. The brain is essential piece that determines what sensors they need.
Adcock assembled Figure AI's team from people with decades of experience, and they come from Boston Dynamics, Tesla, Google DeepMind, and Apple. And clearly has an eye for talent. They are getting close to 100 employees; if you have the skills, do not hesitate to apply. Finding top-notch talents is not going to be solved until AGI is available, so chop-chop, humanoids are much needed!
Judging by how slow those companies have been in terms of robotics innovation and Brett's latest comment on X, it is believable that he picked the best apples from the bunch. I have successfully shipped and sold startups a few times; news like the one below, particularly in the context of automakers, is very hard to come by.
They loved the idea; there was zero friction in getting them on board and doing work together. It took nine months to get them to agree to the press release and several weeks to have them internally agree where it was okay to put their logo on some web pages where our logo was in the surroundings of it.
I fully feel the pain of making this major deal happen and Brett's cautiousness in disclosing other deals in the making.
In researching this piece, I found that way too many are using the expression, "This guy is the next Musk". I don't know personally, Brett, I look forward to finding an opportunity because his passion and work inspire me, but if I was him, I would be flattered and pissed at the same time.
Even if the space for robotics can be easily evaluated in the trillions of opportunity size, Figure AI and Tesla are competitors with different styles and clearly deep different leaders' personalities. I like them both; however, given our conviction for humanoids to become an integral part of our society, Adcock's super focus on that ONE-THING gets my love bias off the bat. He is also following Musk's model in getting noticed without spending ad money!
To the sense of urgency, one of the people in the known I spoke to, put it in this way:
I think Musk got a whiff that Adcock was going to announce something about the bot and wanted to give a sense of "we are making progress, people," before figure was going to show off. So he posted their bot folding a shirt. Later, they remarked that it was teleoperated. So, it was learning. Figure's bot had _already learned how to make coffee on its own. Brett is focused. He doesn't have to deal with the load that Musk has and that plays greatly in favor of Figure.
In the video posted on X, the Tesla bot is teleoperated. That means someone with a VR-set, like, was controlling the robot. A common practice for producing a dataset from which the neural nets improve trajectory and forecasting of the robot movements. Basically, it is how AI can learn and self-evolve within the range of what it learns.
Figure AI was a step ahead of that learning process and, by all appearances, by several weeks. Figure 01, the name of their first bot, was already making coffee on its own.
A neural net system had already acquired that knowledge and was a barista in the making. An incredible feat, not a general purpose operational move as of yet, given that the bot learned from that specific coffee machine, pod, and a specific radius of where the cup was going to be placed, but nonetheless, a foundational piece that clearly indicates, with more data the same framework can teach the robot to be more generic at doing useful things.
Adcock, just like Bill, Elon, in their respective areas is aiming to the one thing that matters the most in the future of humanoids business: RaaS.
You make the PC & software (to hold the initial analogy), and then you license it to manufacture first and then homes. In terms of revenues, nothing compounds faster than a subscription model!
You can make anything "as a service" the difference, in value, is driven by desire of that service. And I have really hard time to accept that an elderly person wants to deal with maintenance and updates of bots, that companies wants to own the lifecycle and have on stuff people that know very closely how to operate their intricacies. Owning anything comes with a degree of responsibilities that makes the pie less appealing from the legal point of view if you are a company or a lot of complexity if you are just a consumer. It will become like an iPhone but that will take what will feel the longest decade in human history! And that is if we know when we can start counting that hypothetical period of time...
I wanted to shed some light on the founder's background before going after the robot details; for a company that is set to impact the planet profoundly, it's important to leave traces behind about the person who might activate Skynet while making coffee. 🤣
It's a marvelous piece of engineering. It's not done yet; as a matter of fact, there are ZERO humanoids currently in the market or nearly close to being operational, but nonetheless, what can be derived so far from Figure 01 public information shows an intense focus on building THE general purpose machine.
|7 Degrees of Freedom (DoF)
|3 Degrees of Freedom (DoF)
|6 Degrees of Freedom (DoF)
|0 Degrees of Freedom (DoF)
|6 Degrees of Freedom (DoF)
|41 Degrees of Freedom
|Torque-controlled walking for real-world navigation
|Race helmet-inspired with an LCD screen for communication
There's a total of 43 actuators and a few nuances that indicate how much thinking there's placed in their design choices. The hand has three knuckles instead of two. When your energy consumption is measured like the Apollo of return mission, a nano detail like that indicates that there are intentional design choices for the application they envision the robot doing.
The Figure 01 AI bot stands at a height of 5 feet, 6 inches and has a base weight of approximately 132 pounds (60 kilograms). It also has a total payload capacity of 45 pounds (20 kilograms).
Based on video analysis, the bot appears to move at a speed of 1.3 meters per second. Granted that camera, compression, and all other usual factors are at play, but as a rule of thumb, I feel good that is the current (in development) speed. Please, correct me in the comments, if you have more detailed information.
At this time they are trying to finalize the hardware details, the speed of movements (walking and turning) will become a focus, on the software end, once they can rely on the speed of decoding what's comes from the sensors and what they can match from the 3D models they ingest in their visual decoding.
There are sensors in the feet which seems to be augmenting the algorithm for the the balance. As I mentioned in our Telegram channel, don't confuse bipedal with humanoids on wheels, they are not the same thing.
To put it in a crude analogy. If there was ever a Titanic for humanoids, the one with wheels we would be taking the lowest deck of the ship and the bipedal would be the ones that the lifeboats were for. They are both robots, just on a different scale of complexity, energy consumption and trajectory calculation when the bot operates in space. If your Roomba had feet instead than wheels it would not choke on wires, humping the carpet edges or slow down on pet hairs, so there's that...
Someone our Facebook page asked "what's your favorite feature of Figure", I answered with just one image. I LOVE THAT DESIGN!
In the next article about this amazing humanoid, I will dig more in what the company is doing to set the standard for "useful work done a humanoid that a customer would pay for".
Below there are some notes that I took during my research and I plan to reach out to Brett to gather something that can fill the blanks on the topics below. For the time being, stay cool, coz the bot's are coming! I can't wait until I can have a Figure 01 on the property doing useful work.
- Challenges in developing humanoid robots:
- Making robots walk bipedally and be stable
- Creating dexterous hands and fingers that can manipulate objects
- Developing AI that can understand and respond to the physical world
- Why now is the right time for humanoid robots:
- Advancements in computer vision and AI have made it possible to create robots that can learn and adapt.
- There is a growing need for automation in many industries.
- Figure AI's future plans:
- Start deploying robots in real-world applications with customers in 2024.
- Triple the size of the AI team over the next year.