At this stage, all that's missing is the data. Figure 03 made it onto the cover of Time's Best Inventions list, and the CEO has spoken.

Has Figure, which recently launched its third - generation robot, Figure 03, and captured widespread attention, once again sparked heated discussions due to the CEO's remarks?

Here's what happened...

Some netizens tweeted that Figure CEO Brett Adcock recently gave an interview to Time. Time reported that Figure 03 still has significant problems and is not yet suitable for daily work.

When asked what else was needed, Brett replied: Data.

If you think carefully about what he said, you'll find that what he ultimately meant was that scale is the key to making robot technology useful in all work fields, and this is another area where big hyperscalers can profit.

The statement caused quite a stir.

Some netizens straightforwardly said there was nothing wrong with this statement, "Data is the new oil."

However, there were also many opposing voices. Some netizens retorted:

I disagree. But of course, he has more insights. In my opinion, the problem lies in the lack of a reasonable architecture and computing power. Humans don't need a vast amount of data to perform simple tasks.

Amid the controversy, some netizens also believed:

Obviously, this is why they avoid on - site demonstrations. But there's no denying that Figure is doing groundbreaking work on the autonomy of humanoid robots.

Ultimately, this debate caught the attention of Brett Adcock himself.

Unexpectedly, Brett Adcock came forward to refute it, stating bluntly that "this is not the case."

Most of the operation strategies we demonstrated to Time achieved high - performance operation when dealing with extremely challenging tasks in a home scenario. Currently, the bottleneck is indeed that Helix lacks data, but I wouldn't describe the problem the way you did. We will definitely solve it.

So, what exactly did Brett Adcock say? Here is the translated and organized content.

"Data can solve almost all current problems"

Brett Adcock: Figure mainly designs and manufactures humanoid robots. Our goal is to enable robots to do as many things as humans can in the real world.

Actually, we often think of our work as "giving a body to AI." This might sound a bit strange.

Ultimately, we're thinking about how to make robots enter the world, be useful in the business field, and eventually enter every household around the world to do things like washing clothes, washing dishes, and cleaning rooms...

Over time, we hope that humanoid robots can gradually achieve what humans can do in the real world.

Q: I saw the robot folding towels and putting them into the washing machine. It's obvious that it can do these things well, but sometimes one of the towels would fall out of the washing machine, and it couldn't pick it up. How can we move from this stage to the ideal general robotic overlord you envision?

Brett Adcock: Overlord? Well, okay. I think simply put, we believe that data can solve almost all current problems.

Moreover, we've seen this in almost all our demonstrations. For example, in the early stages, whether it was folding towels or applications in the logistics scenario, the robot didn't perform well and made frequent mistakes.

But as we input more data, the robot's operation speed increased, and the number of errors decreased. Although it still makes occasional mistakes, the error rate has dropped significantly, almost approaching zero, close to the human level. For instance, when we let the robot run continuously for several hours, its performance was very stable.

Brett Adcock: One of the core focuses of our development roadmap is how to make robots eventually enter every consumer's home around the world.

We hope that robots can take on the things people do at home every day, especially the things I personally hate. I don't want to do laundry, wash dishes, or clean the room.

We haven't made any significant breakthroughs in home automation for decades. So in our view, robots should not only be companions in the home, capable of storing deep memories and communicating with you via voice, but also eventually be able to do all the things you don't want to do at home.

Brett Adcock: Keeping robots extremely safe in the home for a long time is a very difficult problem.

Currently, we're focusing on about 15 safety areas. First, physical safety issues like system safety engineering need to be done very well. We also need to do a good job in cybersecurity, such as how to implement large - scale encryption?

Our safety director previously worked mainly on mobile robots, and the cybersecurity director is from one of the top cybersecurity teams in the United States. They each lead their teams to carry out work.

We're now investing a lot of effort in researching how to safely enter American households first and then gradually expand globally while ensuring that the robots can complete tasks efficiently.

Brett Adcock: It's like manufacturing "synthetic humans," right?

These low - cost "synthetic humans" can work 20 to 24 hours a day, 7 days a week, and always stay in a state of work and learning. How many of these low - cost robots can be sold around the world to help people?

I mean, I think the potential labor force could be in the billions. I think everyone would want one. So, the global total demand is close to 10 billion units.

I really think that one day, when you walk out in the Bay Area, you'll see more humanoid robots than humans. It's incredible to think about.

I've read a lot of crazy science - fiction novels since I was a kid and have always dreamed of such a future.

Now, I think it would be amazing if humanoid robots were everywhere, if you could talk to them just like you talk to people, and if they could help you with things like a partner and even understand you better than anyone else. So I'm really looking forward to such a future.

We design robots not to harm humans or be weaponized. Our goal is to make robots do truly useful work for humans.

One More Thing

Time magazine recently released its list of the best inventions of 2025, featuring 300 annual innovative achievements from around the world. Figure 03 made it to the cover.

It's worth mentioning that Marc Benioff, one of the current owners of Time magazine, is better known as the CEO, chairman, and co - founder of cloud - service giant Salesforce.

And Salesforce also participated in Figure's latest billion - dollar financing round.

In addition, products from Unitree, DeepSeek, Huawei, BYD, etc., also made the list.

Reference Links:

[1]https://www.youtube.com/watch?v=4ZP943 - gARQ

[2]https://x.com/kimmonismus/status/1976581803457810464

[3]https://x.com/Figure_robot/status/1967936969482342891

[4]https://time.com/collections/best - inventions - 2025/

This article is from the WeChat official account "QbitAI", author: Xifeng. It is published by 36Kr with authorization.