I learned more about robotics from listening to Bernt than I have throughout my life. He is convincing that usable robots are in our near future, both in homes and industry.

Several weeks ago I saw one of your videos about the Tesla Bot. In this video you stated that no one except Tesla was developing humanoid bots using neural networks instead of heuristic coding. That same same week I saw a video showing dozens and dozens of clips of humanoid robots being developed all over the world - Europe, China, Japan, and Israel. Many were probably still using heuristics but some of them were using AI neural networks. I am glad you are now looking outside the bubble and have been interviewing many companies and experts to discover what the rest of the world is doing. Many are using AI and end-to-end neural networks. Many are designing the physical bot for mass manufacturing. Some are teaming with experienced mass manufacturers (Samsung, Hyundai, etc.). We should be careful with the use of surpurlatives without extensive research of the global market.

@andrasbiro3007

Ай бұрын

I don't know what was the date of the video, but it was true for some time. I think it took like a year for other humanoid robot startups to come out of the woodwork.

@GET2222

Ай бұрын

The problem with these small startups is manufacturing. 100 droids is nothing. This will be their problem. Musk knows this. He wants 1000s of TESLABOTS in The Tesla factories working 24/7. The data flywheel will surpass these guys quickly. Him building 1 million bots anytime soon will never happen.

Best conversation on humanoid helpers I’ve listened to!

Tendons: by chosing the material that the tendon is made of you can very easily change how compliant it is. Things like steel cable and dynema cord have almost no stretch after preload and things like nylon line and rubber bands have hudge amounts of stretch. There are many materials to chose from each with diffrent stretch, strength, wear, friction, etc. properties.

A great conversation. Love the clear and even philosophical explanations of Bernt.

@orion789

Ай бұрын

That's Europeans.... as opposed to, now, Americans.

@mhfs61

Ай бұрын

that’s why it probably speaks to me, as I’m European as well.

This young man is fulls of good insights. Impressive.

Bernt is clearly a genius!!

Burnt is super-duper smart. So impressed. He's a difference-maker for sure!

My favorite 4 takeaways. 1. developed for the Home, not another factory robot. 2. way cheaper then 25k (a 10k maid robot would sell very well). 3. In customers homes in 2025. 4. Neo far more important then Eve, because we constantly use our limbs for counter balance.

Exciting. Just a guy, one who talks an awful lot like a robot…is getting my heart pumping! This video is archived. As a treasure to my collection. Pleasure, being Bernt😅😅

Excellent discussion. I'm 80 yo and would really find use for a robot.

Such a relief not having John and Scott talking over each other. I was able to learn things from this interview

Herbert-This was your BEST episode ever! You have become a MASTER INTERVIEWER. Now, let's find out about these soft expanding tendons used to power a 35 lb robot to be able to lift 70 lb objects. . . . and by the way I WANT ONE!

This guy is as insightful and as well rounded as James Douma, would be interesting to get the two of them together. And of course Walter Scott. This was an amazingly informative discussion.

Great interview! So interesting and exciting that Bernt is approaching H-bots from a very different technical and market direction. Seems like a brilliant guy. However successfully deploying and supporting one-off robots to non-technical consumers seems to me an unlikely business model at this juncture. AI is not yet ready for that level of variability and uncertainty. Hope I'm wrong and 1X is a smashing success.

With bots building bots, the marginal cost becomes intrinsic cost of components plus energy required to create parts and integrate parts. That tends to the intrinsic cost of manufactured parts, which tends to intrinsic cost of materials at extreme volumes.

TESLA needs to BUY THIS COMPANY! 1X what a perfect fit!

This is in line with what James Douma said recently, namely the software isn’t the problem (or it won’t be very soon). The problem is building many many millions of reliable bots are a good price point. 1X is doing well but we are still essentially looking at prototypes. It might seem impressive that they’ve build 100 bots but they’ve been going for 9 years and they are making a bot that doesn’t walk. I still don’t see any other bot maker who is likely to do a better job that Tesla of scaling up to mass manufacturing.

@corwinzelazney5312

Ай бұрын

Bernt literally said his NEO at home is capable of going to the fridge and bringing him a beer. It walks.

@ArendJanV

Ай бұрын

@@corwinzelazney5312is there a video of that? This guy sounds a bit to much like Trevor Milton to me.

GREAT INTERVIEW GUYS!!!!!!!!!!!

The bot market is starting to divide into different target segments. There's the 180 pound super human brute in the new Atlas, which looks like it's going after jobs that are too heavy for a human, along with the 146 pound Agility Digit, which is targeting highly repetitive container movement. Then we have the middleweight bots with full fingers for more delicate, lighter workloads in Tesla's 103 pound Optimus and Figure's 01 at 121 pounds. 1X appears to have the lightweight category to itself at under 70 pounds for safe home use, but we have yet to see their next gen bot that is supposed to have full hands and legs instead of wheels. Let's hope they can get it working and still keep to that light weight.

Tesla has an advantage in real world vision data.

@brunoheggli2888

Ай бұрын

No

@donp11

Ай бұрын

@@brunoheggli2888 Millions of cars collecting data every second they are being driven. No one else has this yet.

@brunoheggli2888

Ай бұрын

​@@donp11Others will do that to and they are selling way more cars!

@BryansUploads

Ай бұрын

​@@brunoheggli2888Tesla does have the real world data advantage, millions of vehichles gathering data. Tesla is no longer compute constrained. A simple Google or AI search will tell you who has the data lead. Research this and you can find out for yourself.

@Botoburst

Ай бұрын

@@brunoheggli2888 No

As a a robotics engineer I have too say 1X is the most likely to win this humanoid race. The more I learn about them, the more I realize how underrated they are.

@yahanaashaqua

Ай бұрын

Hard to say... engineers aren't publishing updates everyday! I believe it's going to even out and they will all pretty much have the same basic capabilities. What it really comes down to is manufacturing, who can scale the fastest?

@Botoburst

Ай бұрын

@@yahanaashaqua Yep, sounds like they'll be the lucid of bots.

@whowhy9023

Ай бұрын

Extremely likely to be one of the winners in humanoids. Tesla’s bots will not be good in a home.

@whowhy9023

Ай бұрын

@@Botoburstnah Lucid is toast. These guys are very good. They understand cost. Lucid don’t understand economics.

@Roguescienceguy

Ай бұрын

​@@yahanaashaquaeasy answer. It's always the Chinese. They even have a very important incentive because of their aging workforce.

1X for sure looks like a great horse to bet on!

Wow! This may be the most informative video I have seen 😊😊

Very good info here, with lots of common sense, and logical data.👍 The 'humanoid' form factor in robotics is relative to the required function, and the human environment. Most factory, security, and tedious/repetitive work does not require the human form or the same datasets. Android/drone 'commercialization' will be cheaper and faster to market than the interpersonal human/humanoid 'relationship'. Yes, ADT can already provide security through intelligent use of small UAVs and Droids, cheaper and ready to deploy. The human factor in production/productivity will be 'rapidly' declining with mass industrial robots, who will change the entire paradigm of human labor.

The soft-cushioned outer cover all over the bot and its incredibly light weight is for safety in the home!!!!!!!!!

Anyone else notice he sounds like a young Elon Musk? He sees the big problems but breaks them down into smaller problems without losing sight of the end goal.

Optimus looked to be all alone until recently. Now, there’s an explosion of competitors. That needs to be factored into investors’ models. Same will happen with utility scale battery backup.

@Maustah

Ай бұрын

As long there is demand, all those companmies should be able to fill there pockets in the first 5 years. Not every robot compnmay might get to mass production

@andrasbiro3007

Ай бұрын

Maybe, but it's not guaranteed. There is a ton of competition in electric cars too, yet Tesla is comfortably dominating the market (outside of China), and it's true competition is still gas cars. Just because someone decides to compete, has good ideas, and works hard, success is far from guaranteed. A lot of thing has to come together. And even then Tesla has a lot of big advantages, like having far more resources of all kinds than anyone else.

@okklidokkli

Ай бұрын

@@andrasbiro3007 Optimus alone??? Tesla came late in the game, many years after 1X.

@andrasbiro3007

28 күн бұрын

@@okklidokkli Tesla was the first to announce a humanoid robot. AFAIK. And everyone laughed at the very concept. But if they came before Tesla that's much worse. Then their progress is much slower than I originally thought. In less than 2 years Tesla caught up and left them in the dust. That tends to happen. Anyway, I didn't write anything about Optimus being alone. There will be multiple winners, because the market is way too big for just one solution. But very likely Tesla will be the dominant player.

Best robotics presentation on U-tube. Thanks to all 3 of you. Also it explains Elton's recent moves. Clear the deck to focus on Optimas. I think he has seen the competition.

If I were a robot, this guy would be my best friend.😢

1:08:08 I think a way to let the robots be tele operated when failing at a task and create a openai branded platform where humans can login in Vr and teleoperate a neo and complete tasks and use that data to train bots in the largest data flywheel ever ! Vision Pro 2 + open Ai “remote task completion” in 2025 ! This could work !!

Superb interview and guest Hubert. Well done, thanks

simple concept: this is the "Michelin Droid" soft and safe and pliant, but what else does it have except strong partners? THAT may be a lot, of course, but everyone is getting that.

how can we invest??

The most insightful video on bots I've seen to date! Well done guys!

Problem with fuzzy safe robots is that they can't be used in hospitals to replace care workers in lifting and moving patients. That to me is the greatest need for automation anywhere in today's world.

39:20 translation: "Your robot will play when it has no work to do."

Some epic quotes. Great conversation.

This company seems to be going the right way about it.

This is amazing. I never thought of any of this.

The cost of Optimus at 25 K is at customer price, not construction cost.

Looks cool to me. Market for all sorts of bots!! Just like there are plenty of phones in the market

I Doubt 1X will beat Tesla 🤣

Let’s wait and see

I love a good competition!

Trying for the $$ with click bait title.

@dennisFarmer1993

Ай бұрын

You don't know the meaning of!!!😅😅😅

@whowhy9023

Ай бұрын

It’s a good factual title. These guys are good. Above all they are sensible, can’t say that about Elon… Elon does not do low hanging fruit. He likes pain.

@0ooTheMAXXoo0

Ай бұрын

They already have robots deployed in commercial use. They have already beaten Tesla to that step...

@fubutthole

Ай бұрын

Most of his titles and thumbnails are clickbait. I still like the channel.

@Withnail1969

Ай бұрын

Wait for tomorrow. I'm free, free falling

John, we probably can do 10k, 100k, or 1,000k simulations faster than we can do 100 real world data collection examples because of the speed of the computer simulations.

@MachineManGabb

Ай бұрын

Quality over quantity, it doesnt make sense to have 100000 in sim demonstrations if they are not representative of real world physics. Having real world data for real world performance is necessary.

@captcurthress is not a fanboy. Just read some of his other comments

@luckx6692

Ай бұрын

Fan isn’t someone being a fan only when stock prices rise!

Tesla set them off with end to end on a mass manufacturing scale , ie fsd.

Way too early to tell. First 10k units to market should give us some clues. The software needs very fast evolution into the usefulness category...

FSD can drive better than the best human drivers because it can have the same skills but faster reaction time, never get distracted, and have a 360 view.

Yes, but can it dust, vacuum and clean windows?

Consumer space is a huge TAM. It will be easy to sell domestic bots.

Great content. Much appreciated. Can’t say the same for the title :(

Will they offer a robot with a mullet?

i worked with a guy who had the same monotonous gravely voice as sam altman...it drove me crazy. also, he used to burn his arm with cigarettes. i said to him once "that must hurt" and he said "It's supposed to."

@RobertLoPinto

Ай бұрын

The term for this voice which Sam Altman has is Vocal Fry.

@romteb

Ай бұрын

This unatural vocal fry reminds me of the forced low voice elizabeth holmes used to speak with, both characters seem to have a very fake persona and psychopatic traits, i wouldn't trust Altman at all.

@mattsparks5957

Ай бұрын

​@@RobertLoPinto Seems to be more than just vocal fry but not a lack of vocal fry.

@wildcatmahone-md6me

Ай бұрын

Was his name Charles Bukowski? 😂

@corwinzelazney5312

Ай бұрын

🤦‍♂️Right, because if someone talks like another person, they must have the exact same psychology. Loon.

Fantastic! Yes, I am of the industrial first mindset, so this is very contrarian for me. Very useful so I can think about how I might be wrong. My view is that the home market will be very, very hard because of customer support issues as well as returns by enthusiastic customers with unrealistic expectations. Steve Jobs was right that the mass market wants stuff that just works and is plug and play. The early home adopters might be able to put up with stuff that would be unacceptable to mass market consumers. Ironically, he may have a point about the robot being like a child causing a noise at night -- parents may be wired for that but the rest of consumers are going to be very unforgiving. But, parents might put up with the noise at night if they got some real relief from constant overload. I could imagine, "You just go a child you need a nannybot!" Ironically, the "Nannybot" would be taking care of the parent when the parent says OMG my child just got these rashes and is having trouble breathing what do I do? And the Nannybot would take that as something like an Alexa query but provide more useful interaction -- perhaps a follow up question and then call the Doctor or call an ambulance. But, cultivating the right expectations in consumers will be hard and failure to do so will create impossible customer support issues that will ruin the company.

During the initial adoption of humanoid robots, there are dozens of bot companies that may be successful operating in the dozens and dozens of niche markets and may never compete head-to-head with Tesla. For example, the new Boston Dynamics extremely flexible robot might find work cases where it may be impossible for humans to accomplish, but highly suited to a humanoid robot with many more degrees of freedom of their joints than the human body.

Another Tesla killer. I’ve seen this story before.

@RandomGuyOnYoutube601

Ай бұрын

Yeah. Just like with Figure, my spider sense is tingling. They are constantly refering to Tesla. They want to outElon Elon. And they even changed their name to 1X. Hopefully this is not another Nikola story.

@singed8853

Ай бұрын

There is nothing to ‘kill’ here. Tesla doesn’t sell a humanoid robot. No one does. Tesla is a zero in profits in this area just like the other companies.

@yahanaashaqua

Ай бұрын

Tesla is in the best position for mass-manufacturing a robot, they've been doing it already for many years

@whowhy9023

Ай бұрын

No you have not. These guys are real. Much more comersial & sensible than Elon. They are taking the low hanging fruit. I love that.

@MarkXHolland

Ай бұрын

​@@whowhy9023More commercial than Elon. Ha! Ha! Ha! 😂

really impressed 1X approach which is similar to tesla in terms of design and manufacturing. what is still bullish is the need for real world data. whoever can build 10s of thousands of robots first will have a real world data

Their bot looks not like bot i would like to have. I like T800 design.

EVOLUTION AGE OF PLENTY

Jetson's future. Rosie the robot is coming.

Wow!!! I want Elon to see this video - a treasure trove of useful insights.

Would this be purchased or rented. Feels like Robot as a Service would be a thing.

may the best win

If the price of hourly human labour drops to $7 like China and South American countries then the western world would have not need to develop bots to replace $30 / $40 per hour commercial laborers. Developing bots to replace human workers was inevitable as demands for higher wages keep increasing. The world cannot afford today hourly wages as inflation makes life harder even for workers themselves.

Sounds like his fridge needs a vacume relief valve.

Tesla has the factory in Texas they will scale faster than cars.

He sounds like Ilya

Does anyone know what the employed droids actually do in the security industry?

Safety is important, but many are taking the issue too far. A stepladder is potentially more dangerous. Thankfully, Elon's promise that you'll be able to outrun Optimus appears to have been sensibly abandoned. Imagine being told that cars would eventually kill 50 million people.

John… please consider skipping these click bait titles… your original approach was way better than what you are doing now imho. 1X, together with Figure reminds me a lot of companies like Nikola and other vaporware companies. I may be wrong - time will tell 😀

@whowhy9023

Ай бұрын

Not even slightly click bait. These guys are good.

John, the title “Tesla will never get there” is only the half of the sentence 1X CEO Bernt had made. The rest of it goes: “… if they continue manufacturing this way, but they iterate quickly.” which makes this a positive sentence indeed. I like you and appreciate your knowledge, but do you need to use these click bates?

Farmers will be the first to buy Tesla bots if they can be trained to harvest

A humanoid robot being useful is all about the brain. If Tesla figures out the brain they can make endless body types, including one that’s lighter and softer. So really 1X has virtually no advantage over Tesla

Kinetic Energy = Velocity squared Learn something every day. Substitute “Data” for the word energy and I get it, The Parallel Equation, it gives us a bearing on where real-world learning is headed, as important as E=Mass times C squared. Mankind, learning which was is cosmically UP, …and then there is I, seeing the starting point as depicted in the scene of 2001: A screaming ape throws a bone into the sky, some tens of thousands years ago…wandering into The Dawn of reasoning intelligence. Yes, things are moving fast.

@markmarco2880

Ай бұрын

I’ve got to hand it to the guy ,putting in the first word as Safety into his first-principles business model. We are talking about the sum total of power, with which comes enormous, relative responsibility. God, but I love this show.🦬

Better not take until 2036!

so 1X will beat TSLA? I am confused.

Something tells me these guys cannot scale. To dismiss Tesla bots as a far reach is a red flag. They know a bit about scaling.. hands on experience. These guys (and everyone else) don’t. End of story. We make our own coils without mentioning actuators is another red flag.. just saying

If you are making a "Humanoid Robot" it takes from 10 to 20 years ,,, about the same it takes to develop an adolescent or an Adult Human. The Type robot you get depends on the years of training and the Lessons you will teach it..... and there are BILLIONS of HUMANS being trained and Educated every day. The Olympic Champions and Porfessional Sucessful ones are not very common. It will take OLYMPIC LEVEL trainers and coaches as well as Eienstein level teachers to TRAIN the BEST HUMANS not to mention the Froidian, Cecil B Demil and Robert Redford Wisperers to round out their minds and personal abilities to perform their tasks!!! Coping with Life's Tasks and dealing with the requirements of Human sociotie's demands will require All that to be the best that they can be and be successful!!! Doesn't it??? or maybe 1/2 that time if you train 18 or 24 hours a day!!! Lets just hope that the Trainers are not militairistic or power motivated to make Dangerous or Evil bots...

@TM-vs4wt

28 күн бұрын

robots don't need to eat, rest or sleep . you can train them 24/7

👍🏽👍🏽

Human workers building robots, and those robots being sold, is a losing game. Robots building robots, and those robots building more robots is how Elon will win. Tesla will not sell robots. Optimus is a Tesla factory worker making more robots, making robotaxis, making semi trucks, making batteries, making cars, making everything Tesla sells.

Ngl I’d never consider their bot just because of its goofy appearance….

@jonathannumer5415

Ай бұрын

That is a factor

@corwinzelazney5312

Ай бұрын

If you're talking about EVE yes. But this is about NEO, that looks much different and not at all goofy.

@fractalelf7760

Ай бұрын

@@corwinzelazney5312 Neo looks like a carjacker, even worse…

Tesla stock price under $ 100 within a year.

No feet is not real robot , period

What i love is all these small, med, large companies and startups are only coming for Tesla. Its never the other way around. I’ll take the 800lb gorilla anyday over all of them!

At this point tesla bot seems to be behind in terms of both software and hardware as compared to tne competition. Elon has openly said he is not comfortable making tesla AI power if he does not have 25% shares. I don't think elon anymore care about us tesla retail investors.

@bigdougscommentary5719

Ай бұрын

So sell ur stock

@aviralsinghal1274

Ай бұрын

@bigdougscommentary5719 i am and have been a long term investor. And as an elon and tesla fanboy i want them to succeed. But right now its painful to see how tesla is going forward.

@creditrepairwizards

Ай бұрын

Elon did the tesla car thing and is bored with it , he doesnt care about it any more as its been done. NEXT! Now hes going to move on and put electrods in people brains or what ever . Know that this is who he is, these are juts toys to him, and he puts them aside when hes done playing with them, know when to get the hell out, befroe he gets bored and walks out the door and does soemthing else

@churblefurbles

Ай бұрын

ah yes, behind in the "useful labor" of being a security camera drone.

@andrasbiro3007

Ай бұрын

1. No, Tesla is still far ahead. You just don't see it directly, because they aren't showing anything. I can't even list all the advantages Tesla has, that pretty much guarantees their lead. For the hardware you can see Optimus is the most refined, and it was designed for mass production from the start. Also Tesla has been hiring for the Optimus production line for a long time now. As for the brains, Tesla has been working on real world robots for many years, long before anyone else. A self-driving car is a robot on wheels. So Tesla has far more experience, a gigantic data pipeline, more compute than anyone except Meta, and so on. 2. Elon would not stop developing anything just because he's a bit uncomfortable. Development will go full tilt until he's no longer confident he can get the control he wants. The proof is the robotaxi prioritization. Everyone thinks it's still crazy, I'm pretty sure there's pushback even from the Tesla leadership. Maybe that's we two execs left. Sound like Elon who isn't comfortable? 3. Elon cares about investors, not traders. Tesla is always a long term stock, you can't expect the price going up every day. And it's true in general, no matter how well a company is doing, it may take years for the market to recognize it. That's actually the edge we retail investors have, we can see these a mile away and buy up the shares before Wall Street starts the feeding frenzy. Recent example is Nvidia, if you followed AI, you should have understood what's going to happen short after ChatGPT launched. Tesla just had it's "ChatGPT moment", FSD is now working. This is exactly the time to load the boat, because there's little risk and gigantic upside. 4. Also, Tesla's main issue now is not Elon, but the economy. High interest rates, and uncertain economic conditions are very bad for the car industry.

terrible sound from the guest

Your click-bait headlines are ridiculous because no one knows how well any humanoid robot will do at this stage. This approach completely undermines any credibility that you may have had in this space. You turn off way more listeners than you attract!

Your click bait titles really turn me off. If you continue I’ll probably stop watching your channel.

@vasthefox

Ай бұрын

Most smart KZreadrs do click bait titles because it's what's necessary for the algorithm and clicks n views. You may stop watching but it makes sense for the health of the channel. If you can't deal with this reality then channel loses a viewer but survives to be available to other viewers.

@paulmcgreevy3011

Ай бұрын

Ok bye

@JMeyer-qj1pv

Ай бұрын

I think google needs to revise their algorithm to punish click bait titles that don't match the video. It should be possible to do this using AI. I too am getting tired of seeing click bait titles and was sorry to see this channel start to use that approach. But if you think this channel is bad, there's a good AI channel by Wes Roth where every title contains words like shocking, stunning, insane, etc. and the video never lives up to it. Lots of youtubers are of the attitude if you can't beat 'em, join 'em, so really you should blame Google for causing this trend.

@alsetalokinalset

Ай бұрын

Well said 👍

Sad to see the clickbait title...but I guess that's what feeds the algorithm :/

Wow! Sounds like a great company and the CEO isn't a right wing nut!!

Wheels make everything so much simpler & cheaper.

lol. i didnt see one working robot in this vid. garbage but nice try

I don't know if anyone has noticed but Elon isn't a genius and talks out of his ass quite a lot.

I had trouble following the speaking and meaning of the words spoken. Awkwardly presented.

openai stopped working with these guys that tells me all i need to know about them. likely vaporware

They will not beat Tesla. Their first real robot will only be capable of 2 - 4 hrs on a single charge.

I have more faith in Elon.