Why Robotics Data is 250,000 Years Behind LLMs

LLMs have is like 250,000 years of data.

Oh my God.

Um, and meanwhile, in robotics, we have like two years of continuous data.

Yeah.

Like, it's a tiny bubble. And Robotics 2.0 for me is a definition of, uh, sort of, some sort of generalization where you can literally have a robot capable of doing any task in any environment.

Mm-hmm.

And you as an operator, you don't even need to program it. You just tell the robot to do something.

Mm-hmm.

And this is where we are heading. We're gonna have more free time. We need to prepare for a future where robots are gonna take- [upbeat music]

Welcome to the Nebius for Startups podcast. My name is Evan Helba. I'm the head of physical AI at Nebius, and I have the pleasure of being here with Lukas Ziegler, who's a very prominent thought leader, uh, investor, advisor in the space, and we're excited to dive into all kinds of topics, trends, themes, and big ideas that he's seeing out there in the world of physical AI. So Lukas, thanks for being here.

Man, thanks for re- inviting me over. It's great to be here.

Of course. Of course. So let's start off with a little bit about you and your story, and how you stumbled into this unique role where you get to wear all kinds of hats and-

Yeah

... really just, I feel like you just play in the robotic space. You play and you advise and you invest. So talk to us about what you do now and then how you got to where you are.

Awesome. So I'm, I'm used to, you know, sharing robotic stories. Finally, I can share my, my own story.

Mm-hmm.

Um, so yeah, it's not super sexy. So I've been in robotics for the past 10 years.

Okay.

I started as a software developer, then I went through project management roles, commercial roles, business development, sales, um, ending as an advisor. And then I was also posting a lot on LinkedIn.

Mm-hmm.

And at some point, like the big brands from the physical AI space just were like, "Lukas, maybe you want to collaborate at some point, do some content for us, for the brand awareness-

Mm-hmm

... and so more people know about our products-

Mm-hmm

... about our vision, how we want to change certain verticals."

Mm-hmm.

I was like, "Yeah, let's do it." It was still part-time.

Mm-hmm.

Um, so I was working for startups and in my free time, I w-

Mm-hmm

... I would be traveling all around the world to cover some robotic stories, and that's what I'm doing for the past three years. So I made a whole journey from working for different robotics companies in different roles. Right now I'm just a robotics evangelist, I would say.

Mm-hmm.

The one missing muscle I think is, uh, how to help companies get capital. That's why I also try to grow some muscles in that space.

Mm-hmm.

Being an angel investor, scout and advisor to c-

Mm-hmm

... to, to, to startups and young companies.

Mm-hmm.

I actually prefer to call them young companies.

Mm-hmm.

They are doing serious stuff.

Mm-hmm. Of course.

Serious effort. So that's what I do right now.

Nice.

Invest, create content, make companies more viral.

So I've heard you talk about this idea that we're in the era of ro- of Robotics 2.0.

Yeah.

Can you help define what that means?

Yeah, sure.

And contrast it to 1.0.

Absolutely. So let's, let's just, you know, fasten the seat belts. We