We'll Build 100,000 Humanoids by 2027 | Bernt Bornich

0:00 Intro
1:16 Why humanoids?
5:40 AI transfer of knowledge
8:40 Power of scale
9:50 Heritage of Eve
11:50 Tendron drives
18:35 Letting robots experiment
22:00 Designing for safety
25:07 Intelligence diversity
27:45 Moravec’s paradox
32:50 Difficulty of reasoning
35:40 Training models & data
40:40 Humanoid ecosystem
42:50 You can’t retrofit cost
46:30 10x every year
56:30 Hire for who you are
57:50 Be nice
1:00:00 Trust your intuition
1:05:50 2040
1:07:50 Bernt’s favorite Sci-Fi
1:09:09 Individualism
1:12:00 Life philosophy

[jb_kc__]

after watching a lot of Brett Adcock interviews (whose company is doing the same thing) this dude is a breath of air. goes into the technicals, makes you buy into his vision, and clearly shows a lifelong passion for the space

[davidmetzler]

This is so cool. It seems so obvious that this is gonna have a huge impact. The only question is how long it’s gonna take.

[UnexpectedMaker]

That was a really great interview - So many thought provoking ideas and topics. Well done!

[benmcreynolds]

I love Advancements more than most people. I just wish our society could adapt so we can utilize advancements to the fullest. We Require everyone to afford our cost of living. Yet we don't consider how unobtainable it's becoming. This is a common feeling here in America recently. It's like no matter what you do, it doesn't matter, you can't afford to even pay your month to month bills, you feel worthless, it's getting so out of control.

[johannesdolch]

He says a lot of interesting and smart things. He is also very charming. But we also have to remember that prototypes are easy, production is hard. And I'd be really interested to learn about their Compute cluster and data pipeline. These things are not easy to get right.

[WillysGarageNorway]

We need more humanoid robot episodes 😀😀 this is a very interesting topic and thanks for your episodes on Neo

[numgun]

Awesome interview, I'm glad its norwegians leading the race in making the best robots possible, not china nor wall street oligarchs. Hopefully this tech will benefit all of humanity and not just a few rich smoothbrain thugs.

[aadarshktofficial]

Really good interview I have watched in a while and completed it.

[Shin_Ookami]

Really amazing video. Love how you are getting into enough detail to get a deep view of how all of this might play out and what minds are behind 1X but not derail too much into 5 hour philosophical podcast type of interview (although I would probably be interested in that too.) But what left me thinking was his opinion on purpose and not being able to be happy without a purpose. Because I have kind of an opposite opinion on that. Especially with the future he is describing, I think there will be less and less humans who actually will have to have a meaningful impact on humanity because everything will probably be done by robots and AI utilized by a few very bright minds. With automation on the way it is now I think humanity should find a way to be happy with not having an impact on anything, otherwise we will hit a gigantic wave of depression and suicide (a little extreme but you get the idea). Would love to have a discussion on that topic with him.

[injukyoshi]

Like a Kaiju movie, a lot of these presentations, with some reason, are getting a chorus of "GUY IN SUIT! GUY IN SUIT!'

[bcus]

How cool is that idea about training the robot to imitate a recording of a human!

[yoyo-jcqg]

joaquin phoenix did a good job asking questions 👍

[davidallyn]

I like where this company is headed, and I hope they are successful, and I love the vision of a "robot in every house". However, I'm hearing a disconnect between the academic nature of "can we do it" and the practicality of the product. I find it hard to envision selling 1M Neos, which would mean 1 out of every 100 households in the U.S. would have a Neo in the next few years. Personally, I would not spend even $5K on something that would do what I do - even if it was the best robot for feeding the cat or taking out the garbage for example. But I could see a farmer buying a $20K or $10K industrial version of Neo to do general, hard-labor chores in the barn (e.g. cleaning stalls, feeding livestock, "These hay bales need to go up in the loft", etc...) autonomously. In those applications, there might be a market for 100K Neos, the price would need to be marketable, and the robot would need to be trained on doing dirty, dangerous work.

[meowththatsright]

Make a female version if you want a lot of money

[Techtalk]

Theres like 15 humanoid robots being worked on right now. What makes neo or 1x better than the rest?

[ArtByLukeW]

This is such a great video. I love his life philosophy.

[eswyatt]

I believe in tendons AND antagonistic pairs! That's how we get compliance and prescision in biology.

[JoshFlorii]

Great interview! Better mics would be great!

[KarandeepJhand]

This is straight up building AI companions like in the fictional Halo Universe.

We teach the humanoids to ignite the outward exploration of the human experience by kickstarting a new space race. In the process the AI that runs the robot meets a standard of reasoning consistently higher than their human companion. Once the imprint is defined, it then translates that to an exo skeleton for the real human, with added variables like, zero gravity and a system that retains oxygen and pressure.

All of a sudden you got better fitting space suits, with AI Motor synergy for optimal performance.

Its one step closer to Spartan Suits.
At that point you watch out for civil wars and then aliens.

[ondrejcernobila]

28:51 The Mary the Superscientist Paradox has a simple answer, assuming that experience is determined by the physical state of the brain. If Mary, learning everything about colors, leads to a physical state of the brain that occurs when seeing colors, then she will experience the familiar sensation when she leaves the black and white room. If her brain has never been in the physical state that corresponds to seeing color, then she will learn something new.