Can 54-year-old Elon Musk compete with the 39-year-old rising star in the robotics field?

Can you believe it? Elon Musk is being upstaged by a young man.

In 2025, the two major players in the humanoid robot field - Elon Musk of Tesla and Brett Adcock, the founder of Figure - announced almost simultaneously that they would launch their third - generation robots. However, when Figure 03 made its debut on October 10th and quickly went viral, Musk's Optimus remained "partially hidden". He only released a few clips of the golden - shelled robot demonstrating with a popcorn machine and emphasized that it was only V2.5.

The more obvious gap is also reflected in their specific narratives about the new products:

Also on the X platform, Brett neither talked about algorithms nor showed structural parameters. Instead, he simply let the robot appear in a real family space, gently place dishes, stack pillows, avoid pets, and pick up a notebook.

These actions were smooth and natural, filmed in one take. What the audience saw was not code, but a tangible sense of "life" - as if "the robot could really work at home". In this way, Figure completed a transformation from a "laboratory product" to a "visible sample of future life".

In contrast to Brett, Musk still focused on engineering logic: AI chips, actuators, and the supply chain. These are crucial for R & D and investors but difficult to translate into stories that the public can intuitively perceive. So, when Figure captured the public's attention, Optimus seemed more like an internal upgrade rather than a real "debut".

Of course, Musk was still confident at the earnings conference, saying that Optimus V3 was progressing smoothly, would go into mass production in early 2026, and might become Tesla's greatest product. However, in comparison, Brett is better at making the future "visible", while Musk is still concentrating on "building it".

This competition is essentially about two ways of realizing the future: Brett makes the robot enter the public eye first, while Musk is polishing the underlying system.

In today's era where "attention is king", whoever can capture the public's mind first has a better chance of defining the future.

At 39, Brett is turning Figure 03 into a "future sample" that everyone can see; while at 54, Musk may be facing a more pointed question: When the future needs to be seen, will he be half a step behind?

When the new - school meets the old - school in humanoid robot making

Tesla and Figure entered the humanoid robot race almost at the same time.

In August 2021, Musk first announced at Tesla AI Day that he would build a humanoid robot - Tesla Bot (later renamed Optimus) - capable of undertaking "dangerous, repetitive, and boring tasks". The vision he described was that in a world built for humans, the robot could walk safely and assist with daily work. The media quickly dubbed this plan "the ultimate form of Tesla's AI strategy" - to give artificial intelligence a real body and let it enter the real world.

Five months later, Brett founded Figure AI in San Jose, California. With the first sum of savings from selling his electric - aircraft startup, he took a leap of faith to start the robot project.

One of them is a "system - builder" from the self - driving empire, and the other is a "machine - builder" who just completed a career transition. At almost the same moment, they chose to embark on the same unknown path.

From 2022 to 2023, the two companies have been keeping pace in the development of humanoid robots.

In October 2022, Tesla launched the prototype Bumble C, which could walk and grasp with both arms, though it was still an immature body.

A few months later, Figure 01 made its debut - a company that had only been established for 24 months managed to make the robot take its first step.

At that time, the outside world was more inclined to view Figure as a "latecomer" to Tesla. Brett once mentioned in an interview that Figure was "standing on the shoulders of giants like Tesla and Boston Dynamics" and "learning a lot from their practices." But he added, "We just have to be faster."

The real turning point came in the spring of 2024.

Figure 02 was released and integrated with the OpenAI model. In the popular video, the robot tidied up the kitchen countertop and handed an apple - it was not a show but a task. It could understand voice commands, recognize objects, and plan actions.

At that moment, the "humanoid robot" became more relatable to the public for the first time.

It was also from that moment that the landscape of this field was redrawn:

The new - school companies represented by Figure choose to awaken the machine with language and make it "understand humans"; while old - school players like Tesla and Boston Dynamics still focus on polishing a reliable body with steel and algorithms.

In 2025, this divergence became even more obvious.

Figure 03 was released, and its self - developed Helix system made its debut. Brett said in public, "Helix doesn't make the robot smarter, but enables it to move from'seeing' to 'acting'."

The fingertips of this generation of robots can sense a pressure of 3 grams, which is enough to hold a paper clip; a small camera is embedded in its palm, allowing it to complete close - range recognition even when the main camera is blocked; the outer layer is covered with multi - density foam and flexible fabric to prevent pressure when in contact with humans.

More importantly, Figure started building factories.

The large - scale humanoid robot production factory BotQ was inaugurated in California, with an annual production capacity of 12,000 units. The production of robots shifted from "manual assembly" to "assembly - line manufacturing".

The humanoid robot dataset factory Project Go - Big was launched simultaneously, collecting daily task data from thousands of Figure 03 robots - from the kitchen to the warehouse, from the living room to the office - to build a "memory bank of the world" for the robots.

This is a typical mindset of an AI company: let the intelligence run first and then improve the body.

Elon Musk chose to go in the opposite direction.

At the beginning of 2025, Tesla announced the closure of the Dojo training center, integrated the AI5 and AI6 chip architectures, and invested all resources in computing power and control systems.

The Optimus team also abandoned motion - capture training and switched to pure video learning - letting the robot "watch" how humans complete tasks and then imitate.

Musk believes that vision is enough to be the gateway to intelligence. The advantage of this method is scalability, but the cost is time - consuming.

Video can teach the robot to "see" but not to "touch"; it knows "what to do" but can't feel "how much force to use".

Therefore, every mistake has to be tested on a real machine.

This may also explain why Musk's Optimus V3 has not been publicly demonstrated for a long time.

It's not procrastination but the price of caution - while Figure is performing in public, Tesla is still tuning the engine behind the scenes.

By the autumn of 2025, the two ends of this timeline had clearly diverged. Figure 03 convinced the world that robots could "naturally" integrate into life with a single - take video of a family scenario; while Optimus V3 only left behind that short clip of operating a popcorn machine.

Figure 03 made the world believe that robots could "naturally" live among us with a single - take family - scenario video; while Optimus V3 only left the short clip of operating a popcorn machine.

Both companies are chasing the same goal: to make the machine learn to live. The difference is that Brett makes people see the future first, while Musk is still polishing the reality that leads to the future.

The "cooling - off" Musk and the "accelerating" Brett

You can still see the same "excitement" in Brett Adcock that Musk had when he first started making robots. The difference is that Brett's popularity is still rising, while Musk's passion has started to cool down.

At the 2024 Tesla shareholders' meeting, Musk once again painted a grand picture of the future: "The ratio of humanoid robots to humans could be at least 2 to 1 and will definitely reach 1 to 1." In other words, he envisioned that there would be 10 to 30 billion robots on this planet, and the annual production rate could reach 1 billion units.

"Even if Tesla only accounts for 10%, that's still 100 million Optimus robots per year."

He even set a specific target price - $20,000 per unit, which means that once mass - produced, Tesla could dominate a market even larger than the automotive market.

This kind of grand narrative used to be Musk's forte, but in the past year, his tone has become more restrained.

In 2023, he firmly stated in multiple public speeches that "by the end of the year, thousands of Optimus robots will be put into use in Tesla factories."

However, in the summer 2025 conference call, he simply said, "The prototype is now walking around in the office." Even in the recently announced Q3 conference call, Musk's statement was rather conservative. He said that Optimus could "walk around and guide visitors 24/7" at the engineering headquarters in Palo Alto.

The once - exciting tone about the "imminent arrival of the humanoid robot era" has quietly disappeared.

This change is not a matter of attitude but the weight of reality - problems with the robot's hand, issues within the team, and problems with the pace of development.

Musk admitted in a program called The All - in Podcast that the hand of the humanoid robot is the most complex and tricky part of the entire system.

He said, "The human hand has evolved to be extremely complex. It can perform both high - intensity and delicate operations simultaneously, and it's almost impossible to replicate this multi - degree - of - freedom coordination with existing structures." "Most of the hand muscles are actually in the forearm and are driven by tendons." Musk described in the program that the manufacturing difficulty of the robot hand is "between that of the Model X and the Starship." In the latest conference call, Musk mentioned again that "making a hand as dexterous and capable as a human hand is the most difficult part of the entire robot."

This dilemma is more directly reflected in the supply chain - even if they are willing to spend money, they can't buy suitable actuators. According to Tech Stock, Tesla currently has a large number of Optimus bodies without forearms and hands in stock, and the production of these key components is "still seriously behind schedule." This means that Musk's blueprint for "mass production" still remains at the laboratory level.

Beyond technology, the instability of the team has also affected the progress of the project.

In June 2025, Milan Kovač, the vice - president of engineering and the person in charge of the Optimus project, left the company. Immediately afterwards, Tesla sued a robot startup founded by former employees, accusing it of stealing trade secrets related to the manufacturing of robot hands.

The instability of the organization has further slowed down the progress - Tesla had set an internal goal of producing 5,000 to 10,000 Optimus robots in 2025, but the plan has been postponed. Musk said in the latest conference call, "We will be ready to showcase a production - intent prototype in the first quarter of next year, probably in February or March. We will start building a million - unit Optimus production line around the end of next year."

Musk's concern about "control" has also more or less affected the pace of Optimus development. In the latest conference call, he said bluntly, "If I build a large army of robots, will there come a day when I'm excluded from those in control?"

Meanwhile, Brett and Figure AI are advancing rapidly.

Figure's acceleration is not just about the achievements like "Helix" and "BotQ" themselves, but about the fact that it has simultaneously connected the three lines of R & D, production, and data within a year.

Brett has emphasized in multiple interviews that Figure is an "only focus company" - it has no other business lines and no asset burdens.

This means that it can bet almost all its resources on one thing: to make humanoid robots enter the real world.

He said in an interview with Forbes, "We're not building robots for show; we want the robots to truly enter the world."

This statement reflects the thinking of emerging AI companies - first make it "seem to work" and then make it "actually work" in reality.

This is the sharp contrast between the two:

Musk's approach is to steadily build a "system" - to make the robot truly controllable, manufacturable, and reusable; while Brett's approach is to make the "intelligence" visible first - to capture the public's imagination with videos, models, and data loops.

Their divergence corresponds to their ages and positions.

As the CEO of Tesla, Musk has to leave room for error in Tesla's overall AI architecture, energy network, and chip system. He can't let an unfinished robot become a risk to the stock price.

On the other hand, Brett doesn't have such a burden. He can bet all his resources and narrative on one story: to make the robot enter human life.

Huaxin Securities pointed out in its research report in the first half of 2025: "Musk's work focus is returning to the robot business, but the parallel development of multiple lines has led to a temporary lag in the progress of new products."

Musk used to be good at "t