Musk seems to have fallen for the idea that since he's had great success in one or two areas, he can easily repeat that success in other areas, even if the difficulty is much higher and entirely different skills are required.
SpaceX and Tesla are great achievements, but they are primarily stories of successfully implementing and improving existing technologies with innovative engineering and business practices (like iterative design for rockets), to capture markets that were ripe for disruption because they were controlled by a few large players who had no real incentive to improve.
Problems like level 5 self-driving and humanoid robots, in contrast, are different. We don't see these things because they are hard problems that require basic technological breakthroughs that may not yet exist. Self-driving is one of those things that's relatively easy to get 95% of the way to, but that last 5% is nearly impossible. And useful humanoid robots are probably a generation or two away from being possible; there are just all sorts of fundamental issues like energy storage, object manipulation, locomotion, etc, that are all so far beyond the current state-of-the-art that they're almost science fiction.
I don't see this happening in any useful form in the next 5-10 years.
I feel people said the same thing about him for SpaceX and Tesla back when he was just the wacky PayPal guy. Remember SpaceX's first launch was in 2006 (before the iPhone existed!) and it's only started getting really serious traction in the past 5 years. Elon's no stranger to things that don't have any "useful form" for well over a decade. I'm quite certain he's aware of the realistic timelines, and he sells a different story in the press for the sake of fundraising and PR.
I think he's really got figured it out tbh - as long as he expects the technology to exist at some point, the fact that he's parked a big chunk of cash in the idea gives him an advantage. It's like he's reserved a spot for his company to exist when the tech exists. His team doesn't necessarily have to be the one actually making breakthrough innovations in order to be in a good position to reap the advantages of said innovations.
That is a good point; Musk does sometimes announce things that essentially serve as a placeholder for "we're going to do this thing once the technology improves enough to make it feasible", and this is likely more of the same.
But "useful humanoid robot" is way more forward-looking than anything else he's done this with; the closest thing would be the whole "full self-driving without LIDAR" thing, which keeps getting pushed back and is likely not possible with current technology. And building a humanoid robot that can actually do stuff is probably an order of magnitude more difficult than that.
If I had to make a bet on when it was possible, I'd say 25 years. Maybe 10 if Tesla goes all-in on robots and spends tens of billions on it, Manhattan project style.
> If I had to make a bet on when it was possible, I'd say 25 years. Maybe 10 if Tesla goes all-in on robots and spends tens of billions on it, Manhattan project style.
I have a decade of experience in robotics (including both humanoid robots and self-driving cars) and I think this is optimistic.
Robotics is a wonderful field if you like to work on hard research problems. It is a terrible field if you're trying to invent a profitable product.
Why? I mean, what are current problems with making this work?
Hardware/mechanical part? Software? Energy source? All of those? I would assume it's software, same as with level5 self driving (?)
No, robotics is not a software problem. If it was then it would actually be feasible. All it has to do to be useful is walk around without bumping into things, Boston dynamics already has the software for that.
Industrial robotics focuses on single task, highly specialised machines that have few constraints on form, size etc. This is not that.
What tesla are conceptually pitching here (let's ignore that they'll deliver something that's pretty useless by comparison to what they're hoping you're assuming they mean) is a not-in-our-lifetimes leap in robotics in multiple ways
That 100% something else. Industrial robots are almost allways in a cage for safety. I know there are 'cobots' now, but I just never saw one, except in showcases. Industrial robots just do a fixed task and are dumber than a slice bread. Industrial robots are super heavy for what they can lift. Even they are very efficient with energy, they are still allways on main grid / no batteries.
The huge task is anyway not to build a robot. The huge part is to make the robot (or any device) smarter than a slice bread.
True, but humanoid robots have little to do with industrial applications. I'm not bashing Tesla or Musk in any way, but this seems more of a mix of PR and research. Robotics is indeed the future, there will be huge demand in a few years, and of course they want to keep a foot in that door, so they need a project which today allows them to say "we're doing robotics too", tomorrow could evolve in big "ooohhh!" and in the meantime let them build experience in the field. In due time all money spent in this field is going to return.
In my opinion this robot will probably evolve either into a companion aid for lonely/ill people or maybe a future Tesla luxury car chaffeur, rather than doing any serious mechanical work. For most uses, today a wheeled/tracked platform with grabbers is still cheaper, stronger, better. The knowledge accumulated during the journey to get there however will represent the most valuable asset.
that Tesla is in the process of building up expertise in. The Fremont, Nevada, China, and German Tesla factories are all filled to the brim with industrial robots.
Musk isn’t actually looking for profitable products. He’s trying to be the king of vaporware. Remember the Cybertruck? Or how he said he could control pigs by putting microchips in their brains? Or how Tesla was all in on Bitcoin until it wasn’t?
He probably doesn’t even do it to bump the stock price. It’s probably just about doing whatever will impress Grimes for a little while. Then they both smoke so much weed they forget the previous never-gonna-happen idea and come up with a brand new one.
This is just such nonsense, sometimes I question if people ever actually adsorb information or just form lose association and then form an opinion based on how much they like the person involved.
> Remember the Cybertruck?
Cybertruck is literally not vapoware. It was never even supposed to come out before the end of this year. Technically they are not even late yet. Now they have pushed it back into 2022 but that is a few month delay is really not a that big a deal for a truck product with a new architecture based on a new factory.
The way you talk is as if the Cybertruck was supposed to be out last year and was delayed to 2025.
Literally every indication is that this will be a very successful product for Tesla. It being a few month, or half a year late is really not that big deal for a engineering and manufacturing project of that scale.
> Or how he said he could control pigs by putting microchips in their brains?
That was never what they actually said. Why the need to just make up stuff? And again, the never claimed this was a product anybody can just go out and buy within a year or so.
Neurolink the company is still working on that and literally nobody expected them to have a product yet.
Why do you feel the need to shit on the engineers that work on this just because you have some issue with Musk? The companies first goal is to make a product the help people with different deceases and disabilities. Do you get equally dismissive of all medical companies that don't have a product in the market after a few years of fundamental research?
So just because you have made up a wild story about Neurolink being about controlling sheep and you don't remember seeing a video of a remote controlled sheep on twitter you conclude that it is all bullshit.
> It’s probably just about doing whatever will impress Grimes for a little while.
This sort of psychoanalysis is literally the lowest form of comment you can find on Hacker News. You should be ashamed of yourself.
> Then they both smoke so much weed they forget the previous never-gonna-happen idea and come up with a brand new one.
Yeah because a Musk company has never actually done what they said they would. Literally all Musk does is make up stuff and that is why his companies constantly fail and never produce any product.
No matter SpaceX will launch 50-80% of total payload to Orbit this year. No matter that SpaceX produced the best rocket engine in human history. No matter SpaceX is universally accepted as the globally dominate space company. They control more satellites then all other government and companies combined.
No matter that Tesla is universally acknowledged as the global leader in BEV, auto-selling the second best most prolific BEV company by 2x and volume and far more then that in revenue. No matter they produced the fastest production car in human history. No matter that they literally designed a new battery and battery factory from the ground up.
No matter that both company are growing widely fast and Tesla being very profitable. SpaceX was very profitable but huge investment in Starship and Starlink likely make it unprofitable right now, but the unit economics are still great on their core business.
But of course, these are clear signs that all he does is sit around and get high and comes up with ideas to impress his girlfriend. Honestly, I can not even imagine what level of dislike or jealousy somebody must have to be so dismissive of somebody else.
If you want to call him a social inept asshole that is fine, but your argument simply doesn't hold water.
Teslas catch fire. Some catch fire while parked. Some catch fire while driving around residential neighborhoods on autopilot with nobody inside. They're garbage cars. And they're not profitable. The company's profit comes from carbon credits. They're basically a subsidy laundering system for all the other manufacturers whose cars are actually profitable.
And I notice you gave me a 6,000-word rant in response to my post, yet you dropped that whole Bitcoin thing as if I hadn't even said it.
people can downvote this all they want, but if you’ve ever had a friend who smoked too much weed, or fallen into the habit yourself, you’ve seen the same behavior that Musk exhibits. too many ideas, not enough follow-through, and the ideas tend to be wild and unrealistic —- he’s a textbook case.
No it isn't. That literally the dumbest thing I have read in this thread. So your friend is the CEO two of the most successful tech companies of the decade.
Specially a person with the history of having the exact follow-through you criticize him for. Have you literally not been around for the last 2 decades?
Funny that this person that you have diagnosed drug addiction is widely more successful then you are.
Elon is willing to commit a lot of capital and effort into his ideas. the production capabilities of Tesla are improving at a significant clip, and incorporating many state of the art techniques. Elon's approach to innovation is a lot of hype, but there is a lot of hustle to it too. I think he would consider it a waste and not fun any other way.
But "useful humanoid robot" is way more forward-looking than anything else he's done this with
I think flipping a giant rocket on its way back down to have it land upright sounded just as nutty as this awhile back, let alone a startup competing (successfully) against the military industrial complex.
The entire car industry is also adopting electric. No, I think Musk has enough notable wins to get a pass here.
It seems as though a lot of people (not the GP) really want to be apart of the narrative change on Musk, whatever humanity’s obsession is with building up and then destroying someone, followed with the redemption story.
He really doesn’t have to prove anything anymore in my book.
But if you really think about it, flipping rocket is a trivial toy problem in simulation. But useful humanoid robot is very hard even in simulation. Just look at some of the SOTA control theory research. Their simulated robots looks stupid.
Universal Healthcare is also a toy problem, but a true albatross concept beunricratically in the US. I say it’s a toy problem because we literally have copy and paste templates from most of the western world that we can use.
Flipping that rocket and getting NASA contracts to go actually do it against Boeing/Lockheed/etc is not a toy problem.
I don’t want to say it would never have happened without Spacex, but probably it would not have happened in our lifetime.
Flipping rockets is a simpler problem in the sense that it the required technology already existed, there just wasn't anyone with enough incentive and resources to actually do it.
Human-level robotics, on the other hand, is not a problem of missing incentive. For all I know about the field (as a casual observer), we are still several breakthroughs away from achieving this goal.
>I think flipping a giant rocket on its way back down to have it land upright
And herein lies the problem: landing the rocket is really cool, but normal people overestimate the difficulty. Because of this feat, "Elon can do anything!"
The reason we don't land rockets is half technical, half economic.
The list of things Elon has said they will achieve, and have not, far exceeds the list of exceptional tech his companies have produced.
I really don't get the problem here. They want to work on a very difficult problem, they have admitted it is difficult.
Are people angry that Musk said he will have a prototype in a 18 month? Are they angry because he dared even mention wanting to work on that problem.
> landing the rocket is really cool, but normal people overestimate the difficulty.
Building the biggest rocket in the world, building new engines that are better then any invented in human history, coming up with a complete new way of reentering and doing a never before done landing maneuver and doing that from a space port that was basically a swamp only 1.5 years earlier is quite different then 'landing a rocket'.
Of course it doesn't mean he can magically do anything but his companies have enough engineering credit between them that when they announce to put serious resources behind something we should at least not dismiss them out of hand.
The have fully vertically integrated AI, power electronics, manufacturing robots and batteries both design, testing&validation and manufacturing facilities. Not to mention lots of capital and a CEO who clearly is supportive of the project if its gone make short term money or not.
I for one am interested to see where this is in 2-3 years.
> The list of things Elon has said they will achieve, and have not, far exceeds the list of exceptional tech his companies have produced.
Questionable. If you give an product announcement a 3-4 year lead time for development and commercialization. When you go back over the existence Musk companies, I would suggest that most has actually happened.
Sure they didn't do Red Dragon, propulsive Dragon landing, Roadster and Semi is not out yet FSD clearly didn't go as he hopped. I guess we could come up with some other abandoned and delayed projects. I guess they just delayed Cybertruck for a few more month but that is hardly noteworthy in the automotive industry.
The Starship projects is still remarkably close to the timeline from 2016 (that was considered beyond delusional by industry experts), Raptor and Vacuum Raptor outperform what was announced, Starlink are regularly launching and they are mass producing phase array antennas. Falcon 9 platform reached the 10 relights that was the initial target. The have managed to deliver humans to Station and bring them back (years before the competitor who will get 2 extra billion will be able to).
Tesla despite people clowning Musk about how dumb he is about manufacturing has scaled manufacturing as fast as almost no other manufacturing company ever has and there is no end in sight (new factories coming online as soon as this end of this year). Not to mention that they have achieved industry leading unit margins on that production. They are by far the largest produces of BEVs in the world by volume and by a even larger amount by revenue. They are producing Plaid power train cars and have made the fastest production car in history.
They are in the final stretch of having completed a re-design and re-engineering of a battery manufacturing line and battery. They did this almost fully vertically integrated building much of their own manufacturing equipment. This project alone is a gigantic engineering effort of a scale that is hard to really grasp. The end of this year is very likely when the first actual cars with these batteries will be sold.
I think FSD is by far their most visible and most relevant failure and by far the most talked about on HN, but overall I think its hard to point to any companies that did better in terms of technology development last decade (specially considering that when the decade started these companies were tiny).
> But "useful humanoid robot" is way more forward-looking than anything else he's done this with
Other aspirational goals he has set include a colony of a million people on Mars and a neural lace. I think a humanoid robot is less ambitious than either of these two.
I feel like with Tesla navigational brain technology, they can easily put together a bot that can mow lawns, sweep and mop floors. Pull out the garbage/bring cans back in. Security Sentry observe. Get mail. Walk dog. Deliver packages. I would pay for a bot that takes those chores off my hands.
We already have the technology to send people to Mars. We sent people on round trips to the moon 50 years ago, so a one way ticket to Mars now isn't really that hard. What we don't have is cost effective technology to send enough people and equipment for a permanent colony there. But at this point it's really just working on getting cost down (which Musk has done beautifully so far), we have a pretty good idea of how we'll get there.
On the other hand we just do not have the technology for a full humanoid robot, neither now nor the near future. An AI that can do mundane tasks like ironing and folding clothes is way beyond our grasp, and then there are the issue of limbs and power too.
> We already have the technology to send people to Mars.
Not for live humans we don't. The record for landing on Mars is around 1 ton for >$1B, this is too little for a crew cabin and ignores the life support requirement for months spend in space.
For sure, they need to last longer which means more supplies and more reliability. But we also got 50 years of progress since Apollo 11, and we wouldn't don't need to pack a return vehicle and its propellant.
And, frankly speaking, it is a goal worth pursuing: not just per se, but for all technological advancements it would need to make in order to be successful.
> If I had to make a bet on when it was possible, I'd say 25 years. Maybe 10 if Tesla goes all-in on robots and spends tens of billions on it, Manhattan project style.
"Useful humanoid robot" is easily pared down to "robot toy with a vacuum cleaner attached, and some funny gestures on top of an Alexa-style Internet querying system". And a lot of promises that this time next year it'll be useful, or at least able to climb stairs. That level of technical functionality is already achievable, and the Musk brand solves the commercial problem of "but it doesn't really do very much"
I like Musk, and I've been following his progress since 2011. I always thought he seemed very intelligent and seemed to really understand what he was doing.
However I have to agree with the person you are replying to. There has been the sort of myth going around that making rockets is hard. It used to be hard... 50 years ago. That is literally the technology Musk has been using for the last 20 years: 50 year old technology (and yes that includes reusability). He has just been doing it cheaper, which there was no incentive for anyone to do before because it wasn't a free market.
The reusable stuff is still misunderstood by most people too. Musk is losing money because of the reusability of his rockets. If he had never pursued reusability he would have made far, far more money from his launch business. Only now, after 10 years, are his reusable rockets just starting to break even with the cost of an expendable rocket (after reuse not before). He is pursuing reusability because it is his passion project, not because there is any money to be made from it.
Since Musk, another 4 private companies have reached orbit, or will reach orbit this year (Firefly, Rocket Lab, Relativity Space, and Virgin Orbit). All of them took almost exactly 6 years from founding to orbit, just like Musk did.
Furthermore, if you have been following closely, you would have seen the huge and humiliating mess ups he had with Tesla where he literally flushed billions and billions of dollars down the toilet. Any COO from another company would have had no problems setting up these manufacturing facilities without trouble, but Musk though he was smarter than everyone else and proved it was the other way around when it came to a competitive business like auto manufacturing.
The man has vision, but this whole AI thing is just Musk getting lost in his own marketing.
> There has been the sort of myth going around that making rockets is hard. It used to be hard... 50 years ago.
For something that used to be hard, the very well-funded companies in that area besides SpaceX seem to be having a hell of a time achieving anything even remotely near comparable performance.
> Furthermore, if you have been following closely, you would have seen the huge and humiliating mess ups he had with Tesla where he literally flushed billions and billions of dollars down the toilet.
I have been following it closely since 2012. Not sure where he flushed the money down the toilet? And with that he still grew the company well. Without Elon Tesla would never have become the 666B company that is today.
> Since Musk, another 4 private companies have reached orbit, or will reach orbit this year (Firefly, Rocket Lab, Relativity Space, and Virgin Orbit)
Rocket Lab was founded in 2006 and reached orbit in 2018 (and failed in 2020)?
Virgin Orbit's vehicle LauncherOne was started in 2007 (under Virgin Galactic) began operations in 2021.
And we'll see how Firefly and Relativity Space succeeds with their launches.
>I have been following it closely since 2012. Not sure where he flushed the money down the toilet?
You have been following for that long and don't know the company continues to burn cash? That they said they'd never raise money again, and then raised money about 6 times? That if it wasn't for subsidies and funky accounting there is no profit?
Tesla has had their fair share of hard times. And some mornings have been horrible after having read some Elon tweets xD
When it comes to raising cash, things change, I remember once cash raise happened immediately after their value had increased much in a short interval, so it made sense to raise. And looking back, with the company being worth 666B now, they've done a lot of right calls, but of course some wrong ones as well.
And it's a myth that Tesla doesn't do any profit. For example last quarter Tesla got around 354M in regulatory credits (this is what the anti-tesla crowd likes to point out) but the net income was 1.14B. And besides the regulatory credit should be counted, not like the oil industry hasn't had it's insane amount of subsides over the years.
It totally makes sense, I agree. Just seems odd to say we won't, and then do, repeatedly.
>And it's a myth that Tesla doesn't do any profit. For example last quarter Tesla got around 354M in regulatory credits (this is what the anti-tesla crowd likes to point out) but the net income was 1.14B.
It's only a myth if you don't actually examine the underlying accounting. Regulatory credits, realizing deferred revenue on FSD (which doesn't exist), expanding the accounts payable, cutting R&D, bitcoin speculation...
All of these things affect the income statement in a way that pumps income that may or may not even be sustainable. All so people will say "look, profits!"
Are they marginally profitable now? Maybe, but that's nothing to brag about at a market cap of over half a trillion dollars.
>not like the oil industry hasn't had it's insane amount of subsides over the years.
So, what's your point here? Did I say oil companies deserve subsidies? Am I looking at the financials of Exxon? Being as how Tesla vehicles have tires, plastics, a supply chain...they also benefit from oil subsidies. What does this have to do with anything?
> Are they marginally profitable now? Maybe, but that's nothing to brag about at a market cap of over half a trillion dollars.
Decent to do more than a billion in profit and they are a growth company with Giga Austin and Giga Berlin opening soon. When amazon had a similar market cap they did around 800m in profit. It's ofcourse easy in retrospect to say that amazon was a winner, but of course it was an unknown to investors and everyone placed their bets back then.
And my point with subsidies is that they affect the outcome, as they exist they have an effect on the business, and should be valued as such.
I think this is where you go astray. Reusable rockets aren't old technology. In fact, they were considered so impractical and complicated that none of the big players even tried to develop one.
In fact, even after SpaceX landed their first boosters succesfully, participants of a panel discussion organized by at an American Institute of Aeronautics and Astronautics Forum on Propulsion rejected the idea as being market-unjustified:
> If he had never pursued reusability he would have made far, far more money from his launch business.
Strikes me as short term thinking and not true, also. The plan is to sell tickets at $400,000 each to get to Mars. There doesn't seem to be any way to achieve that without reusability. The money to be made from founding the first Mars colony supplied with reusable rockets completely dwarfs some random launch provider on Earth.
SpaceX never was and never will be a simple launch provider. If that is your lens then you are missing the company goals of 1M people living on Mars.
First, no one denies that reusability is still not profitable. This includes Musk himself and heads of all the other rocket companies who have all made it clear they aren't pursing it because these is simply no business case for it. At best maybe recovering the fairings or engines could be profitable for now.
> The plan is to sell tickets at $400,000 each to get to Mars.
His Starship is set to hold 40 people on the trip to mars. That would mean he earns 16 million per Starship sent to mars. Each Starship takes about 8 launches to fully fuel.
He currently sells 1 regular launch for 50 million each, so your numbers just don't add up. This isn't a money making scheme.
In fact, Musk would be the first person to admit he isn't trying to make money by sending people to Mars. This is his passion project. In fact, the whole reason he created his Starlink business was to pay for his efforts to colonize Mars.
> First, no one denies that reusability is still not profitable. This includes Musk himself and heads of all the other rocket companies who have all made it clear they aren't pursing it because these is simply no business case for it.
What? Here is "Musk himself" saying in a tweet that Falcon 9 only needs to be reusable 3 times to make sense economically.
Oh ok, thanks for that. The last I had heard was that it would take 10 launches to break even. It seems spaceX has been working very hard making refurbishing the boosters cheaper and cheaper over the last few years.
> First, no one denies that reusability is still not profitable. This includes Musk himself and heads of all the other rocket companies who have all made it clear they aren't pursing it because these is simply no business case for it.
It makes no sense for them because they can't offer a low enough price and otherwise create payloads in order to achieve a flight rate.
ULA simply only has one costumer, the US military and very occasionally NASA. They have never really competition in the private market or ever tried to generate their own payloads to increase flight rate.
Arianespace is a incredibly bureaucratic organization where everything has to go threw a political process. It cost them 4 billion to create a pretty average quality large rocket and doing re-usability for them would take 1-2 decades and another 10 billion.
The idea that it is not profitable or more generally wroth it for SpaceX is crazy. SpaceX has mentioned many times that even only using the rockets 2 times already makes sense for them. They are getting 10 reuses out of the Block 5 rockets. The idea that it was not worth the investment is frankly speaking laughable.
Without re-usability things like Starlink would simply not be possible in the first place.
> He currently sells 1 regular launch for 50 million each, so your numbers just don't add up. This isn't a money making scheme.
While I agree that Mars travel is not a money making scheme, those 50 million includes a huge margin and another huge cost for the second stage.
If you assume almost no margin and you assume a equally reusable upper stage, far less operational cost, no fairing recovery cost, less refurbishment cost and so on you can get the number way lower then the current Falcon 9 price. Depending on flight rate cost in the low millions are approachable for LEO launches.
There are other opportunities, dedicate tanker Starship should reduce the number refueling required. Next generation Raptor 2 versions and new ally for the Starship should eventually make it possible deliver a 250 tons reusable. Meaning you can potentially do 4 refueling flights, not 8.
The pure business case for Mars colony still doesn't add up but low enough to make a Mars bases together with NASA totally feasible.
That ticket price is very, very optimistic. With it's 1200 t of fuel, the starship eats 240 t of methane and 960 t of oxygen. The price for methane is around 80 cents per litre, 240 t is 520 m^3 ish, so thats $420 000. Oxygen on the other hand seems to cost $10.6/kg, giving you a number of $12 000 000. Rounds out to 12 million per a single fueling. If the Starship carries 40 people that's just enough to bring the fuel cost of the LEO -> Mars trip down to $300 000. The first trip to space must cost the same amount in Starship fuel only so that's at least $600 000 per person in fuel. And this is all ignoring the Super Heavy, which will probably raise the cost over a million, in fuel alone.
> Musk is losing money because of the reusability of his rockets. If he had never pursued reusability he would have made far, far more money from his launch business. Only now, after 10 years, are his reusable rockets just starting to break even with the cost of an expendable rocket (after reuse not before). He is pursuing reusability because it is his passion project, not because there is any money to be made from it.
Spacex can launch more rockets per year because they don't need to keep building a first stage. That's how they could launch almost 2k Starlink satellites already. If you look at the statistics, most of the launches for 2020 and 2021 have been on reused boosters.
It's organization that's hard, pulling the trigger trusting your decisions are correct, Musk seems to excel at this; more important are the Boards of Directors and leadership hierarchy that trust his process enough, and arguably he's right enough more than he's wrong.
SpaceX created the only full flow engine in history that was actually used in a flight.
And they are well on track on the first fully rapid reusable rocket twice as powerful as the Saturn 5.
> Furthermore, if you have been
following closely, you would have
seen the huge and humiliating
mess ups he had with Tesla where
he literally flushed billions and
billions of dollars down the
toilet.
Hmmm one correction, SpaceX main driver is Gwynne and I doubt SpaceX would have been even close to what it is now if not for her. Musk is there more front face for the public.
Tesla is his child, but so are many other (now obviously failed) projects.
Ah this old Musk-Hater mantra. The 'Gwynne' does everything trope.
Gwynne is an excellent manager and I don't want to say anything bad about her and SpaceX is lucky to have her but the idea that she is 'main driver' is just idiotic. It is counter to everything you read about SpaceX. It is counter to all the experience reports from people within or from outside of SpaceX.
Neither Gwynne nor any of the other original leaders, Mueller or Königsmann seem to agree with the assessment.
This trope seems to mostly come from an time where Tesla was not doing so well in 2017 and everybody that didn't like Elon was saying 'see Tesla is going to the shitter, he doesn't have Gwynne that is the reason SpaceX is successful'.
Tesla of course now is incredibly successful without having a Gwynne and that somehow is just lucky or something.
For a while people also used to say that JB was why Tesla was technically successful, he was the mastermind behind all the engineering at Tesla and without him would fail to innovate.
Gwynne could leave SpaceX tomorrow and SpaceX would continue to do very well. The head of HR for example has been with the company since Falcon 1 days and could take that job and do a good job as well. Elon could likely do much of it himself but it would talk away time he would much rather spend with the engineering team.
>SpaceX main driver is Gwynne and I doubt SpaceX would have been even close to what it is now if not for her. Musk is there more front face for the public.
A substantial portion of the Musk fanbase believe he's the one personally designing rockets and coding the software to land them.
I've never heard a single SpaceX fan suggest that Musk actually writes code any more. I think that is a bit of an exaggerated caricature. He knows how to do it, and he wrote a lot of code in the early days of Zip2, but since then, he has left the actual coding to others. (And, given he hasn't done it in anger in over 20 years, his software development skills are likely rather rusty. Plus, I’ve never heard any suggestions he has skills in avionics software development in particular.)
I think he does actively involve himself in engineering-level discussions and decision-making, both at SpaceX and Tesla. Most CEOs don't do that, but some do – Bill Gates did that a lot when CEO of Microsoft – so it is not unprecedented. Designing something as complex as a car or a launch vehicle (or an operating system or an office suite) is a team effort, it is never going to be all on a single individual, and senior people are inevitably going to delegate a lot of the detailed analysis and design work to their subordinates – but a senior executive who actively involves themselves in engineering decisions can have a real impact on the outcome, for better or worse. He isn't formally trained as an engineer, but he has a degree in physics, and he is obviously a reasonably intelligent and very dedicated guy who has had years of access to some very experienced engineers, so the idea that he has learnt enough aerospace/automotive engineering to actually make a useful contribution as an engineering decision-maker isn't unbelievable.
I'm a bit of a SpaceX fan myself, but I feel like there is a big disconnect between what the average SpaceX fan actually believes about SpaceX, and the caricatures that many SpaceX/Musk critics like to paint. (Of course, there may be the occasional SpaceX fan who takes it way too far, and for whom these caricatures actually become true, but I don't think the vast majority of us are actually like that.)
> as long as he expects the technology to exist at some point, the fact that he's parked a big chunk of cash in the idea gives him an advantage.
It also let's you point to two (major) successes while dismissing the Boring Company, Nuerolink, SolarCity and the Hyperloop, among others that went nowhere. Not that his sucess rate is anything to be embarrassed by, but his announcement alone doesn't lead to me expecting anything to ever be commercially available.
SolarCity was not Musk lead company, he was just an investor. SolarCity was acquired and is part of Tesla and while they had to remove resources from Solar during the manufacturing difficulty and scaling difficulty the solar and home energy part of their business is still growing very well.
I think its fair to say that going into solar was not as successful as Tesla hoped, but neither is it a failure. In the last year the growth of the business line has been finally showing promising growth.
They are starting to work with more outside contractors and they have the cheapest solar product on the market.
> Nuerolink
The company still exists, they are working very hard on a very difficult problem. They are still adding people. They are working on something incredibly hard, has never been done before and will require a lengthy regulatory approval process.
They did not announce a timeline for a product that they failed to hit.
What exactly is your issue with the company?
> Boring Company
They are growing, have an engineering team working on the technology. Their commercial bids are very competitive. They have established test sights in California and Texas. They have made quite a few improvements already and in the publicly available data their commercial bids seem to be competitive.
Apparently its a failure because the first product they ever did didn't meet the your standards or something. The costumer of that product seems to be happy and wants to extend the project. But who cares about happy costumers, what matters if how much people on HN like it.
So how exactly is this a failure? I would be my ass that Boring Company valuation now is significantly higher then when it started. And I bet its already more successful then most startups.
If that is your definition of 'going nowhere' I wish my projects would turn into companies of that size.
> Hyperloop
Maybe go back and actually read what was said at the time. Musk literally said he wasn't gone work on it. So your critic is that project he never said he was gone do didn't get done?
That makes no sense. He literally said 'I posting this paper, maybe somebody wants to work on it, I have no time to do it'.
If these are considered 'failures' now I really have done nothing in my life. I mean I haven't even created a single 100M+ company yet. Seems to me the only one that actually didn't go anywhere is the one that he said he wasn't going to do.
That's my point. You talked about companies that are 15, 5, and 5 years old as "not announcing a timeline they failed to hit" when I claimed they had produced nothing so far. Which, one, is a crazy standard to volunteer with Musk as he regularly fails to hit timelines he announces. But my point was he invests in things for years - decades - without any public progress but also without pulling out investment. The "heads I win, tails I flip again" is what I was commenting on. SolarCity (where, far from being uninvolved, Musk was Chairman of the Board) is older (in Elon involvement) than Tesla; if Tesla is a "success" how little would SolarCity have to do to be a "failure"?
How do you feel about Blue Origin, which Bezos regularly infused billions into before it finally got to its first flight?
The Hyperloop, is something Musk spent over a calendar year and numerous engineering assets working on before he decided to give up and open-source their work. If any of his projects was an unreserved failure, it was the hyperloop.
I don't get why you're so upset about this. I didn't say Musk was horrible. I just pointed out a lot of his projects tend to fail. I explicitly said that his success rate is good. I just also pointed out he has enough misfires that there's no reason to believe his announcements at this stage are going to see commercial availability.
> The Hyperloop, is something Musk spent over a calendar year and numerous engineering assets working on before he decided to give up and open-source their work. If any of his projects was an unreserved failure, it was the hyperloop.
Can you provide me with a source on that? During that time he was full time for Tesla and SpaceX and they were very busy during that time with other projects. So I can't see how this much engineering was put into it.
Also, not doing a project is not the same as failing at doing something.
The result of the investigation was clearly, I think its a good idea but I have no time to do it. Maybe somebody else will.
I think he still hopes to eventually put a Hyperloop type system into a boring tunnel, but he thinks shorter range solution are more important and tunneling technology needs to improve to make it viable.
> SolarCity
I agree it could have been a failure, but also don't know that for sure.
> I don't get why you're so upset about this. I didn't say Musk was horrible. I just pointed out a lot of his projects tend to fail.
I am not upset, I just pointed out for the most part his projects do not fail. You pointed to one project they didn't actually do more then a initial technical evaluation on. And on project that could have failed, but actually didn't.
I think objectively if any person said 'I'm gone do X project' (that was considered incredibly difficult), I can't think of many that would have more credibility then Musk (ie Tesla, SpaceX or a new company).
A humanoid robot is a risky project for sure and it does not seem to be the same level of importance as batteries or reusable rockets, so maybe they will abandoned it at some point.
There's aiming for timeline and realistic timeline (4x expected, and perhaps 4x that too) to which he seems to shares the former; he may not consciously pay attention to the latter though and that may apply a needed pressure on himself to focus to make sure things are moving along as quickly as possible.
It's a moonshot, I think; he's got enough money personally (hundreds of billions) to fund an R&D lab similar to Boston Dynamics, and since it's under Tesla's wing, there is more than enough money to fund R&D into this.
Of course, a lot of that money is all virtual, in owned stocks and market value.
Counter-argument: I remember seeing a talk by Andrej from a year or two ago and wasn't too impressed - it seemed like they had a huge amount of work left to do and put everything together for a reliable self-driving AI.
Everything they showed today, from the vision models to the path planning to labeling to simulation was bleeding edge. And they seem to have managed to put everything together in what appears to be a tremendous engineering effort. I can't see why they can't use a similar playbook for the humanoid robot and increase value over time. The vision framework they used for self-driving is fairly general and can probably adapted to humanoids. The action policy space is less re-usable but I suspect they have clever ideas on how to tackle that.
Musk may be bad at forecasting but he's darn good at selling stuff that half-works and then assembling a team that can execute on his vision.
I'm speaking from my knowledge of the specific technologies they are using - it's fairly close to SOTA. I'm not aware of an apples-to-apples comparison of Tesla's autopilot with a similar system from the competition. Most competitors are using LIDARs and/or running their autonomous vehicles in limited geographical regions. IMO, the state space and action space is significantly more constrained in those cases.
was my initial thought, from the comments in the video, it's clear they didn't think deep enough about Robot application. So I guess the main goal is to make politicians be busy thinking of universal income instead of BS like "naming of the autopilot". By the time government will decide something in this trials autopilot will be 4-5 generation different software they trying to evaluate. Aren't that first signs of singularity?
Haha, and then in 10 years when He announces X you will say: “since he’s had great success in pioneering self landing reusable rockets, starting a colony on Mars, pioneering the EV market, pioneering Robo taxis, making the first commercial humanoid robots, moving traffic underground, adding computers to people’s brains, he has now fallen for the idea that he can easily repeat that success in other areas like X”
This is going to be a good thread to come back to in 10 years. Maybe I’ll be completely wrong and need to admit I was foolishly optimistic in Musk’s ability, however, my guess is the parent comment will be hilarious/absurd in hindsight.
Love this framing, agreed 100%. I find it extremely frustrating when people try to murder ideas and shot down any attempt at building something new. I know that being a cynic is easier but come on, isn't it great that Musk is pushing us closer to the future? I have a lot of problems with many of his actions but it doesn't mean I can't recognize positive ones. /rant
My opinion on Musk changed with the whole crypto pump & dump thing, for the worst ofc, but those public announcements with ambitious goals are key for his achievements and to get the right people to work on those problems.
There's a lot of people here that apparently don't like that him for saying things lightly, and to pose complex problems as trivial problems - but isn't that the driver to break those barriers and challenge that type of barriers such people create.
"Oh no, this is a whole different problem he shouldn't dare to say this!"
Except he should, and if he gets the right people involved with the right motivations and funding they might actually do something. How can they fail? Even if it take them 10 years to do so.
People shouldn't underestimate the power of those bold statements, for example Musk said in a recent interview that Tesla probably has the best material eng. team in the world - and the thing is that they might not be the best, or the smartest, but they sure as hell are contributing to the leap of Tesla over everyone else.
Will people drop from Boston Dynamics to go work for Tesla? Why not, if some don't agree with the status quo, or because they think that things should be done differently and BD isn't willing to pursue that because they're too invested into their own thing.
So this should be celebrated, more people working to solve this is the right path.
Boston Dynamics has made some pretty incredible progress. The very first videos had Atlas and Spot "walking" but really more like controlled tripping. It was cool and impressive, but definitely in line with my idea of what was possible in robotics. The most recent video has Atlas doing parkour. It's skin crawling. It looks fake. It can jump and flip better than I ever could. It is suddenly out of step with what I ever thought a robot would do.
Conventionally, I would agree that this sounds like an insane overreach from Musk, but it feels like the ground is shifting under my feet and the previously preposterous is becoming suddenly possibly with increasing frequency.
Anyway, better that it gets me my groceries than hunts me down through the forest.
Have you even watched the announcement? Has anyone commenting here did?
He clearly said the bot v1 won't do more than carrying simple things and will be super slow and weak. It will only have very basic grasping hands, even likely with 3 fingers.
It will be what grasshopper was to Falcon 9.
And about autonomous driving, Tesla autopilot seems to be almost as good as Waymo, except with pure vision. And I'd bet a great sum Tesla will move faster forward.
which, anyone naysaying bot v1 doesn't see it. Think about moving a dishwasher. Up and down stairs, over and across pathways which are just a little too narrow for the furniture dolly so it keeps riding into the dirt. The target market for v1 is going to be slim, because I bet it won't be cheap. But it will be endlessly useful, and doesn't need to solve really hard general AI grade robotics problems, just some computer vision things, and without the self-driving car crashing penalty.
The question is: how much? Price it too high and the market just isn't there. Furniture dolly's don't have the same cachet as sports cars. I'm hopeful, because the technology is so close.
> Tesla autopilot seems to be almost as good as Waymo
Tesla's system is better, if only because you can buy it right now. Waymo has sold exactly zero of their cars, and they don't even have a plan to do that any time soon.
Yeah then what's the point? I can already order Uber and someone drives me. The main allure of self-driving is that I have my car, that drives me around, without involving anybody else.
Not if it isn't available to buy. For that matter I have a highly advanced general super intelligence. You can't have it. I'm just using it to solve my kid's math homework.
As long as I can rent it to solve the homework of my children which are in your class that's fine.
In fact if you feel comfortable it gets everything right and doesn't kill people in situations where humans don't please consider extending to a larger area of homework until you feel comfortable letting it carry out PHD level research assignments.
> Self-driving is one of those things that's relatively easy to get 95% of the way to, but that last 5% is nearly impossible.
But there are niches for which 95% is just fine. E.g. autonomous trucks driving through the Nevada desert with manual remote-control during the first and last mile. (Like handover from plane autopilot to ATC-driven landing, basically.)
> in any useful form
It doesn't have to be useful; it just has to be better than whatever Toyota is selling, to win long-term robotics development contracts. (That probably won't be contracts for humanoid robots, mind you. The humanoid robot is likely a halo product to sell e.g. robot arms.)
And then you look at some other things he has part-taken in, like Boring company and ask question how is this in anyway better than what we already have, by any metric...
And I won't even go to Hyperloop, which he had sense not to spend money on...
Tesla FSD might end up getting good enough after a hardware upgrade in 5 years or so. The humanoid seems at least a decade or two away.
That said, I truly enjoy that he's investing time and money in these futuristic sci-fi projects rather than trying to optimize the next quarter earnings.
The border between "improving existing technologies with innovative engineering" and "technological breakthroughs" is pretty murky, though.
For example, Raptor engines are the first Full Flow Staged Combustion engines to ever fly and they should be cheap to manufacture on top of that. Orders of magnitude cheaper than other rocket engines, with target price tag lower than a million USD per piece.
By my standards, this is a breakthrough. By your standards, this may just be innovative engineering. We can both argue our positions plausibly.
I can definitely see applications of having at least semireliable robots on Mars or Moon, though. For starters, they need not mind the vacuum and the radiation.
I was skeptical when I saw it, even as a long time fan. But I thought is using all money capital, _human_ capital and trying really hard to "repeat that success in other areas" such a bad thing? It doesn't seem so, but the despair in comments like these sound very smart, i'll give you that.
Your conclusion made me chuckle: 5-10 years isn't very long to get to a mass produced general purpose robot that's useful; and as Elon said they already make all of the components necessary - including aggressively working on AI including for visual systems - so they arguably only have refinement left and edge cases to figure out next - not a small feat but puts them 10+ years ahead of anyone who will try to copy them re: engineering talent, manufacturing, etc.
As Elon has said people are generally terrible at exponentials, and so majority of people can't see or understand the economies of scale and synergy between everything he's doing - and the compounding effect that has on reducing costs and increasing speed of scaling - which then also compound with each other.
For me it seems Musk has ability to hire competent people because he can verify if you are bullshitting him down to physics. This and being bold and rich and persistent is enough for success it seems. Ah, also marketing/vision to get (more) money.
Energy storage is one of the most problematic aspects. These guys will need to be heavy to be useful for more than a couple of minutes. This creates additional challenges (like quickly moving toddlers). I bet early adopters will be people without small kids.
The underlying cause of the problem of the "$200 asprin" isn't a fundamental problem of healthcare, it's a fundamental problem of the wacky health system of the USA which is—I assure you—the envy of no other first world country.
While it is true that the US health system excels along certain narrow metrics (it's particularly well optimised for providing cutting edge care to the affluent) it is poorly optimised for serving the needs of the median citizen. And it's a dumpster fire when it comes to distribution and affordability.
The solution exists, and that's to stop doing the broken thing.
1. Nowhere indicates Elon Musk believes "he can easily repeat that success in other areas."
2. So do you believe you know better (than Musk) about
- the relative difficulties of disruption of auto and space industries vs this humanoid industry and
- Self-driving
to be judgemental?
I would hold my tongue and meanwhile, wait and see.
This type of reasoning simply does not understand the market competition. Once there is a market, the goal is to win the competition, not to hit some technical bar.
Is there a better player in the industry in building humanoid robot? I rarely see any that is better than Tesla. Modern cars are essentially a robot, or the closest thing in tech stack.
Is Musk stretching himself? Of cuz, that's how new things are created...
AFAIK they were avoiding deep learning like the plague for the locomotion. I don't know if something has changed, but that's why the progress was slow.
It is interesting that many of the top comments are as dismissive as: “No wireless. Less space than a Nomad. Lame,”
When you look at Tesla’s talent in: AI + motors + batteries + manufacturing I can’t believe how negative and dismissive the comments are. Can we not have some appreciation for what Tesla HAS achieved vs only focusing on missed ship dates?
The part that I think is going over peoples head is that for so many jobs a robot can be as error prone as FSD is currently today and still be safe to replace a human. You wouldn’t want the robot to control a fork lift but it could replace many jobs in a factory with out creating major risk.
The flaw with full self driving right now is not that it doesn’t work sometimes. Rather it needs to work 99.999999% of the time. This isn’t true for many 5 mile an hour jobs. If they can get the robot to complete a job correctly 99% of the time, at the right price that robot will be printing money for the company that owns it.
I don't dismiss Tesla's successes at all, but they are primarily repackaging and improving technologies that already exist. They didn't need to invent the lithium-ion battery, or the electric motor, or the automobile; all of those things already existed and were available to be adapted to Tesla's desired use.
Humanoid robots, however, are a much harder problem than building a mass-market-ready electric car. In order for someone to bring to market a robot that can perform general household tasks, for instance picking up dirty laundry around the house, putting it the washing machine and dryer, then folding or hanging it up, then they first need to make multiple fundamental several-order-of-magnitude breakthroughs in the fields of machine vision, robotic locomotion, energy storage, artificial general intelligence, etc.
I'm not dismissive because I don't think Tesla is full of brilliant people who can do amazing things. I'm dismissive because building a useful, mass-market humanoid robot would be an achievement of incredible magnitude that's as far outside of the current envelope of technological achievement as the moon landing would have been to the Wright brothers.
> They didn't need to invent the lithium-ion battery, or the electric motor, or the automobile
They do have their own batteries, their own motors, and a pretty unique automobile. I'm not so sure the point stands. If the "they are only using existing technology" line was true, you'd have seen a ton of companies from China, or the traditional automakers, take the market before Tesla had any chance to compete. Somehow Tesla had time to build their own factories and parts from scratch and still beat them...
Having Rosie the robot maid from the Jetsons at home is the vision of the future being sold, but as you point out, that level of technology is very much still the realm of science fiction. If you're watching the other parts of the AI conference this was presented at, self-driving cars are practically here. The reality, of course, is that that actually aren't yet. But you can buy a Tesla today to get an electric car taste of that future, even though it's not actually quite here.
A bipedal furniture dolly, however, is as unsexy as it is possible, using today's technology. It requires none of the impossible breakthroughs, though energy storage seems like it would be a problem. (Quad-copter drones seem to have follow as a control scheme figured out already.)
Doing this under Tesla means the earliest versions don't need to sell at all, they just need to be useful internally. Which I'm sure is why this is coming up in the first place. Think of all the places a forklift can go in a factory vs a human. The forklift is extremely useful, but still limited to flat, paved terrain. A bipedal robot base solves that, enabling Tesla to make more use of available space.
If this bot follows Tesla's strategy for self-driving (which still isn't here), it'll ship with "all the necessary hardware for Rosie the robot maid-mode", and then consumers will have to wait another couple decades for the software to actually catch up.
You've left out manufacturing. Building a car factory + supply chain is quite the challenge.
Tesla had the best and the brightest - I've worked with them before they were acquired by Tesla on a daily basis: https://teslagrohmannautomation.de/en/
and still had trouble scaling. Manufacturing is considered to be one of the hardest things about building cars - electric or not.
Edit: What's with the downvotes? Disagree? Comment below then.
>Building a car factory + supply chain is quite the challenge
And yet Elon announced to the world he was going to have the most automated factory in the world, the machine that builds machines, and that the factory was the actual product. People lapped this up. 2 years later they were hand-assembling cars in a tent.
Now do you see the problem?
>Manufacturing is considered to be one of the hardest things about building cars
Again, you say this as if it's some kind of revelation. It's obvious. The only shock is that Musk and the people he was surrounded with didn't realize it.
They tried to do something quite crazy with manufacturing and it didn't work. Musk has often ignored industry standard and often it works. In this case it didn't and people will now use this forever to show how stupid he is.
However what happened then is that they effectively corrected their mistake, fixed their production, reached their production targets with excellent unit margins.
They learned from the episode and realized they need to reduce part count and made a whole series of innovation and if you look at how their new Model Y are going to be developed in the new factories you see a huge improvements. They developed a new alloy for casting, worked to put largest casting process into production, totally redesigned the battery for manufacturing and massively simplified the battery pack.
To the point where Tesla are quite competitive in terms of weight and ease of manufacturing. They already have industry leading unit margins and some of these innovations will only hit next year.
That part of the story somehow is never mentioned when people say 'hahahahha build cars in a tent'.
>They tried to do something quite crazy with manufacturing and it didn't work.
What did they try to do that's "crazy with manufacturing" that hasn't been done by every other car company on the planet?
>fixed their production, reached their production targets
Have their plants ever reached production capacity?
>To the point where Tesla are quite competitive in terms of weight and ease of manufacturing.
And here we have the point. Tesla are competitive, fantastic. That's a long way from all of the early claims about upending the entire manufacturing industry through factory automation the likes of which the world has never seen. Funny how that part of the narrative disappeared.
>They are not doing so bad it seems
Looks to me like legacy car makers are catching up, no?
I was responding to the comment above mine that didn’t mention manufacturing but pretty much everything else. It might be obvious to you but general populace have no idea about difficulty of manufacturing. Some engineers haven’t got a clue on what goes into building things by millions.
that makes sense. rockets are big dumb objects with incredibly tight engineering tolerances. humanoid robots are anything but big and dumb, and there's no incremental path to success between a bunch of servos wired to a mannequin and having it do the dishes.
> "but they are primarily repackaging and improving technologies that already exist."
Did you mean: they are innovating? IMO there's not enough appreciation for getting to commodification and scaling the existing tech. It's a very difficult problem, from what I understand.
> I'm dismissive because building a useful, mass-market humanoid robot would be an achievement of incredible magnitude
Agreed. But they don't have to start with "mass-market humanoid robot", there are 1000s of niches where then can execute "land and expand" strategy. Same playbook as with Tesla or SpaceX: start with something niche (sport car for rich people/gov contracts), iterate to expand ($30k EV/satellite internet everywhere).
This isn’t “no wireless, less space than a Nomad,” though. This is like RCA announcing an iPod in 1961, reasoning that because they already make magnetic tape and portable electronics, it’s only a matter of time before they’ll be able to fit 1000 songs in your pocket.
This is a good point that most of us have difficulty to grasp.
The innovation speed is actually increasing.
While Tesla is spending a lot of time and effort on their AI (from hardware to software), they might be able to reuse it later for other things, automating more tasks, and at scale. There is a compounding effect.
I don't think Tesla is kidding itself with Optimus, the v1 would be close to useless, and they know it. But it would allow them to learn, build a team of experts and get to the state of the art, to eventually improve it and go beyond. Maybe the main added value at the time would on the software only i.e. mechanically the robot would not be much better than the Boston Robotics of the day. And it could sufficient to actually make a big difference between a Boston Robotics and Tesla Bot ?
I am curious to see what they will end up with. I am not worry about the timeline. I was not expecting a humanoid robot in the next 10years. Tonight announcement is not really changing it, except that now, Musk put a stick on the ground, and we might witness some actual progress.
Exactly. People were saying it was impossible. He announced it to happen in 2018. It will happen 3-4 years later than planned. And people use that to bash him, forgetting the elephant in the room which is that he made the supposedly impossible possible.
I know. I wasn’t clear enough I guess. I should have said « You said exactly the kind of ridiculous argument that haters repeat ». My answer was also sarcastic.
Well, anything that if you do incorrectly someone (or some robot) will likely spot it before the failure gets to a customer. For instance, if you have a robot that picks fruit but 1% of the time the fruit is not ripe or is worm-eaten or something, then you can have a process where the picked fruit are inspected a second time and the bad ones thrown out.
Another case where 1% failure would be acceptable is weeding. Suppose you have a bunch of robots removing invasive English ivy from a forest. Maybe once in a while they'll pull up salal or a fern or something by mistake, but that's fine because those aren't endangered species and in the absence of ivy the ecosystem will quickly recover.
I worked in a factory that was the distribution centre for all the cold and frozen food in every Safeway in the sate. There were a few hundred of us who had to pickup a heavy box from a conveyor belt, turn around, take a step or two and put the heavy box down on a pallet. When the pallet was full someone would come and take it away. Repeat. For 8-10 hours a day, 6 days a week.
Upstairs people were manually putting those boxes on the belts. It destroyed people's bodies, I'm happy I only worked there for a few months each year to put myself through University.
The factory was very high tech and cutting edge, though the humans at each end of the conveyor belts were unavoidable.
We messed up all the time - dropped boxes, did bad stacking and made pallets fall over, crashed pallet jacks and knocked over pallets etc. The failure rate was probably around 5%, I'd guess.
Starting pay was around $30/hr, double that on Sundays, and 2.5 on Public Holidays
A humanoid robot that actually works most of the time will change the world.
This doesn't seem like the kind of problem that would require a humanoid robot to solve though? It's just that even with current state of the art robotics, things like grabbing cardboard boxes (easily crumpled or dropped, mass distribution within may be uneven or unstable, may already be broken or leaking etc) and stacking them into a stable 3D pallet is actually quite hard, so much so that Boston Dynamics managing to do some parts of this in optimal conditions with uniform, lightweight boxes was hailed as a breakthrough:
For many menial tasks, a 1% failure rate (or indeed any failure rate higher than a human) is generally not a problem so long as the economics stack up and safety isn't a concern.
Every manual job I’ve ever done I’ve probably failed more than 1% of every movement. But then I correct. In a car moving at high speeds (focus matters more), I do this less often, but I have been in crashes.
Im 100% sure Tesla can mass-produce the hardware for such robots. But the biggest challenge is in software, And then the FSD debacle comes to mind.
There's major advances in AI needed to get to full self driving and even more advances for a useful autonomous humanoid robot.
But then again, those will not come to fruition just by waiting, someone has to work on this and if Tesla advances the state of the art, I can only welcome them. But I don't expect a Tesla bot in the way Elon poses here in the foreseeable future.
Oh, I'm 100% sure the robot hardware Tesla will make will not look like what is shown in the presentation, because that design makes no sense. There's no room for joints to move etc.
The BD hardware looks cluncky, but it is robust.
If my robot folds my laundry correctly for 99 days and on the 100th day shreds my wardrobe, I don't think that will print money (at least not for very long).
From what I’ve seen FSD makes minor mistakes that at high speeds lead to danger or death. I can’t see a minor mistake in folding laundry leading to the laundry being shredded. Musk said the robots would not be strong (only can lift 5lbs). So unless the robot picked up scissors it couldn’t shred the laundry.
FSD has not been rolled out yet, so you’re attributing to FSD some errors that are actually made by humans. Yeah maybe these humans were confused by marketing to think it was FSD, just as you think it was FSD, but that doesn’t make it FSD.
Tesla is quite literally selling a package called "Full Self-Driving". I don't see how it has not been "rolled out yet" unless you're confusing Tesla's bullshit marketing lingo for level 5 autonomous driving.
A robot that folds laundry would obviously have torque limits while handling clothing, so the prospect of destruction would be minimal.
The error domain of a laundry robot would be misidentifying an unusual item and incorrectly folding it, or incorrectly sorting it. If someone sold a humanoid laundry robot which folded and sorted my clothes correctly 99% of the time, I'd consider that miraculous.
Sure, that would be bad. So it only needs to be good enough to fold correctly for 99 days, and then on the 100th day just do nothing (i.e., detect a problem and hard STOP).
Then a hard reboot will kick it back into line and we're good to go again.
Obviously that's not good enough for driving a car, but it certainly is in all the manual labor jobs I've ever had.
Tesla is not without problems. Poor drive quality. After spending 40K+ for the car, I was cited for not having front number plate. I had to figure out how to tape the frame, spend more time appearing in court. I prefer not doing such things than having automatic s/w update for fine tuning fart noise.
why did they put it in the trunk when all other car dealers fix it and give ? And why do you see 80% of Tesla'a on the road without the front number plates ? Also if you put the number plate it may not reach their advertised 0-100 mile in 3.x seconds ?
Because a lot of people like to leave it off, especially on expensive cars, not exclusive to Tesla. It looks better and the chances of getting a ticket are actually pretty low. There are plenty of people who've driven around without a front plate for many years without getting cited, and it usually just a fix-it ticket without a big penalty. It's not generally because people are ignorant of the law. And it's not required everywhere either.
If Musk was really serious about this, he would start with smaller steps:
0) Finish the tentacle charging robot they demoed in 2015 (roll out to superchargers)
1) Build delivery robot on wheels using FSD (like the Starship Technology's robots). These could be used on hospitals, possibly warehouses, pizza delivery, cleaning streets, painting lanes on roads, ...
2) Build dog-like robot, like Boston Dynamics's spot with a "hand".
...
3+n) Try to build humanoid robot.
Each smaller step would provide Tesla with the experiences needed to take the next step, as well as an actual sellable product.
This is what a sane CEO would do.
I get that Honda and Boston Dynamics have humanoid robots, but they've worked on them for 30 years, and they are not talking about selling them as a generic product.
Musk is not serious about this. Just like the boring company (see Las Vegas Loop as example), he is selling a _sci-fi dream_ to people either as a distraction, to pump the share price, to attract new talent, or to brand himself, Tesla and the other Musk-companies.
As a shareholder, this actually makes me quite concerned about the sanity of the leadership, the distraction it will cause internally in Tesla, and about the fact that they are starting R&D into something new, while there is still plenty of other more reachable products, they have presented but still haven't delivered. Tesla is a vulnerable company, and they are losing focus.
The worst part is, I suspect that he will (need to) present an even more impossible thing in 2022/2023 to keep the hype-train rolling.
Now, where is that charging robot that actually looked like it worked in 2015?
> Musk is not serious about this. Just like the boring company
Quite the claim. As usual we have the 'look at the very first prototype' its not actually perfect therefore its fake and nothing but marketing.
The Boring company is series and they are doing all kinds of things. Saying they are not series is disrespectful to the engineers that work on those machines.
I'm a shareholder and spending some resources figuring out something like humanoid robot to me is useful R&D if its a product or not.
There is no evidence that this is a significant enough distraction to be worried about leadership. They are doing exactly what they should, expanding production, working FSD and battery production working towards releasing Cybertruck and Semi.
I don't see many wanting a tentacle arm, that's a waste of resources.
I'm also heavily invested in Tesla, have been for years I like this alot, it's an exciting future to work towards. Will it take longer than elon says? Probably. But as long as you're moving towards the goal, it's great, one day you'll be there.
I also think building a humanoid robot can interest a lot of talent, and while building this robot they'll learn a lot that can be applied elsewhere.
There's way too much negativity here and way too few exciting product announcements these days, so I loved every part of this.
We are negative, because it seems either Musk doesn't seem to fully respect and appreciate the difficulty and the steps required to build a humanoid robot, or he is lying about what he thinks is possible at the expense of customers, shareholders, and employees.
I honestly don't understand why they don't use their FSD+electic motors+battery tech and create a bunch of products on wheels; farming equipment, robotic lawnmowers, robotic vacuum cleaners, street cleaners, etc.
I think there is so much potential in autonomous farming equipment, and even the robotic cleaners could be an order of magnitude better than current products, if the FSD logic works. But maybe electric farming equipment is less feasible than a humanoid, and, obviously, cleaners are less cool.
It seems like much more feasible, useful, and lower hanging fruit with the tech they already have, compared to this. It could also create confidence in the autonomous capabilities of Tesla's FSD, and it would lead to sellable products within years instead of decades or more.
Good idea with autonomous farming equipment. Although doesn't really sound like it's aligned with Tesla's mission, so maybe some other company could license Tesla's tech and do it. Because it sounds like it should be done!
If anything it gives scope for smart and talented people to work on something that if fails they can easily spin multiple startups from in various attempts to find a route to success.
This happens all over the place and we should be thankful people are either dumb/smart enough to take these risks on big ideas.
Yes and Boeing worked on rockets for 50+ years, General motors worked on cars for 50+ years and so on. I don't get why people seem to have such a limited memory of the past.
Flipped around, but potentially the same takeaways:
Turns out building a true FSD car is almost as hard as a humanoid robot. They've already got a Boston Dynamic's-like robot dog internally, but it sucks at getting around a parking lot.
i.e. set your expectations of the arrival of true FSD on the car to whenever you think Tesla can actually deliver this robot.
I think their trained data and current models for navigating are almost completely useless for a humanoid robot. They might have the awareness tracking for a certain "scale" of objects but a humanoid robot can trip on a half inch ledge a car can ignore and roll over. Let alone all the control systems for making the legs do leg stuff.
Boston Dynamics is already using modern AI technics and that shows on the huge advancements they had in recent years which, being revolutionary robots, are still kinda useless compared to "traditional" industrial automation costs and applicability.
Tesla is as close to being BD as it is close to being Apple and launching the next iPhone killer.
I think you are right. The gating part in both applications is, whether you can create an AI which is "good enough" to get a real understanding of the environment.
I assume this kind of behavior helps with the "Elon" brand surrounding his companies, but it just feels so overdone, unnecessary, and, as another person said in this thread, cringeworthy.
I saw the iPod nano announcement clip this morning, where Steve Jobs has a camera zoom in on his jeans pocket, talks about how the Mini fit in you regular pocket, then pulled the Nano out of his coin pocket. It's a little silly and got a laugh, but it had a purpose. It showed how small the iPod was and felt, maybe not natural, but at least tasteful.
Sure, but the iPod nano was a great product weeks from being in customers' hands. This is a recruiting event for a decade+ project to do something that hasn't been done before.
In five years, Tesla bot version 1.0 will be a Tesla-brand gig-work slave wearing that silly robot suit. In ten years, they'll still be working on the robotic version 2.0, which they will be perpetually promising to release "next year."
That was hilarious, the way they gave a nod to asimo walking up stairs, the perfect cheesy dubstep music, then the cringe dance... what balls haha. If people can't laugh at this then what the hell
I said this in another comment, but how do they get away with it? People are in this very thread making serious comments like Tesla has even a thread of a chance of actually doing this…
I’d love it if Boston Dynamics recreates this dance with Atlas.
Risky, but maybe it would even be good PR to try to match the dance and see the robot fall over just to point out how far they actually have come in 30 years.
The glass was not supposed to be built prove, maybe work on your listing comprehension.
Also, if you throw a ball like that against a normal glass on a truck it might go threw and kill you. The Cybertruck class stopped it and would have saved your live but I guess fuck them for making a saver vehicle.
And the guy in the suit was clearly a joke, unbelievable the hissy fit people are throwing over this.
From a market dominance perspective this is brilliant. Before now Tesla could only sell FSD that doesn't work to Tesla owners. Now they can sell FSD that doesn't work to people with any kind of car.
This won’t be vaporware forever. Seeing the latest from Boston Dynamics makes that abundantly clear. The utility for general purpose robots is immense in difficult environments like the surface of Mars, where O2 wouldn’t be needed. I anticipate a small number of companies making the hardware, and a large, diverse workforce training, supervising, and maintaining android fleets for a variety of tasks. The legal system will have a lot of work to do, along with the general economy. It’s still far out, but not that far.
I see bipedal robots as sort of a PC of robotics. They don’t need to have great AI, personalities, or any sci-fi jazz, just basic collision avoidance and programmability. Lots of spaces and tools are shaped for humans; this means a specialized robot is no linger needed for every task. It doesn’t mean, as Elon suggests, they will be shopping at the grocery store per se but it could mean they could load a laundry bin and carry it downstairs. With adequate programming, it could learn how to fold specific garments and identify flaws. Like video and spacial APIs, not goal seeking.
Yep. I bet there are plenty of programmers out there who would rather spend their evenings programming a robot to do their laundry than do it themselves. :)
I think they could see use in cities for delivery or patrol type jobs.
The stakes are so much lower for a 50kg robot moving at walking pace compared to a car.
Consider a pizza delivery operation. You could imagine having dozens of robots delivering pizzas, and even if you had to have one or two humans driving around town getting robots out of stuck edge cases when they happened, it could still be a net gain.
Consider that Boston Dynamics Spot robot dog is already operating on industrial sites today. Having it stand upright is not too different from a deployability point of view (see the Atlas robot).
I assume if Tesla actually does intend to build them, they will be targeting a price point in the tens of thousands instead of $1M range. That is, priced comparable to a car.
If a robot cost say, $50k, they could seriously be cost effective compared to humans if they don’t have a ton of maintenance costs.
what we also see from boston dynamics is that this kind of tech takes decades to develop. It is an immensely difficult problem that Elon seems to think is easy and doable in a year.
Industrial use will certainly happen. Personal use not so much though. By the time robots are feasible you'd be crazy to have one in your home. Like all new tech these days robots will be leveraged against you at every opportunity. They'll be constantly recording audio and video and reporting the events and contents of your household to companies and to the state.
Maybe the legal changes will happen eventually to prevent that, but I expect Americans will have general purpose robots long before our laws do anything meaningful about surveillance capitalism and the erosion of our Fourth Amendment rights
I'm still befuddled what the industrial applications are.
At some point a car factory had people moving stuff around and screwing things in and out. Now we have arm robots and conveyor belts and other machines which do not resemble humans at all but are obviously way better. We still have humans involved but I think the best way to minimize how often they are on the floor is with better specialized tools.
I feel this is the equivalent of building a robot to be your rickshaw driver instead of just realizing you can build a self driving vehicle with nothing that resembles a human invovled.
I don't think all robots need to be 1:1 replicas of humans but you might be underselling versatility. One robot for every specific task might be a good idea in some places where as in others it might be better to have robots capable of performing several types of tasks giving you the freedom to assign them to different roles as needed. It cuts down on the number of robots (and back up robots) you need to keep available.
A more human form would allow for that versatility and allow for easier interaction with the world and the items within it which we've built with accommodation of our own bodies in mind.
I am generally a fan of the 'do one thing and do it well' type of philosophy though and ideally we'll have a mix of specialized machines and general purpose ones.
I'm sure there are but why would another robot be better?
If the problem is something like the dexterity of a stationary arm machine isn't good enough - I'm having trouble believing the dexterity of a moving humanoid machine would be different at all.
Converting supermarkets to dark stores. Supermarket by day, robot driven fulfillment centre by night. Using humaoids to pushing and fetching items from a human-centric store.
I've been wrong many times before but still have trouble believing this would be more efficient than just having two separate buildings. Even in places where land is very expensive.
How many dark stores would you have to run in a city like Boston to get the same efficiency you could get from a signle warehouse with a similar system to this?: https://www.youtube.com/watch?v=ssZ_8cqfBlE
And if you insist on doing that multi-use concept - it still seems like a robot on wheels and with much different than human arms would be more efficient.
It's quite interesting how people keep betting against Elon. The HN crowd should consist of mostly risk-tolerant personalities, but the majority of comments are negative. Regardless of what one might think of the feasibility, no other company in the world today are as ambitions as SpaceX and Tesla (+ Neuralink). Feels like what Apple used to be during the "one more thing"-era.
Both Tesla and SpaceX has been discounted so many times before, and despite proving time and time again that they will and can deliver, they keep getting discounted.
"It's quite interesting how people keep betting against Elon."
Because he doesn't have superpowers.
Making a car that runs on batteries is a feat, ramping up production of an automobile is a feat, but within reason.
Making a humanoid robot, something that nobody else in the world has any clue how to do, and which Tesla has no background in ... well that's more than a feat.
So a better question - why do people continue to believe this guy when he makes such claims?
> Making a humanoid robot, something that nobody else in the world has any clue how to do, and which Tesla has no background in ... well that's more than a feat.
Boston Dynamics has shown robots capable of dancing. Can Tesla replicate? Maybe not in one year, but for such a project it doesn't really matter if it's late.
I'm betting against him selectively. For example I'm skeptical about Bot, Hyperloop, Neuralink or FSD happening within 5 to 10 years of the original timeline.
Not to take away from their brilliant engineering accomplishments, but SpaceX and Tesla were dealing with technically solved problems. We (as a species) have been to space, and we know how to build electric batteries and cars.
These were at the advanced engineering stage after enormous scientific progress of the last few hundred years and going all the way back to beginning of physics and math.
With AI we are at pre-science stage, hunting mammoths with flint arrowheads and wooden spears.
I mentioned in another comment that I think this has to do with the very human phenomenon of the narrative paradigm shift (a type of M. Night Shamalyn plot twist).
Goes something like this:
The guy that took on the military industrial complex with his small little startup and actually won, with wild whacky ideas like landing rockets upright, and winning NASA contracts against the likes of Boeing and Lockheed, is actually not who you think he is.
He’s not?
Nope, even though he was one of the single strongest drivers of electric vehicles that is pretty much making the rest of the industry shift after the viability of electric was shown with Tesla cars. The truth is, he actually sucks.
Oh, what a plot twist. Wait, does this have a happy ending now that we know he’s a really a bad guy?
And then we fill in the redemption story.
It’s the creepiest thing, and I have no idea why we keep doing it.
I can think of so many people who are radically more accomplished in vastly more interesting and important ways, it feels absurd to start listing because of how many I'd be leaving out by doing so.
Can you not? Like, how about people who have made major academic or philosophical breakthroughs? Major authors? Even within business, or celebrity, Elon doesn't stick out much other than filling a small niche.
The problem with naming them is that then it becomes a discussion on whether the people I've named are impressive. Someone mentioned Einstein and was downvoted. Einstein is very obviously more interesting and accomplished in my mind than Elon, but he's one of many. Turing was 23 when he wrote his thesis on algorithmic computation. There are people with multiple turing awards out there.
If your yard stick is "rich person who made more money" yes, or "CEO who makes memes", or "is a celebrity", Elon is impressive relative to the general population. It's a sad, sad way to measure accomplishments, but sure.
The point is they needed to make multiple significant independent breakthroughs in established fields.
Turing like a lot of pioneers is seriously overrated. When surrounded by low hanging fruit it’s easy to seemingly make a great deal of progress especially when your credited with formalizing existing ideas or making that one last step into usefulness.
Einstein actually pushed boundaries in a well established field which is vastly more impressive. Musk significantly pushing the boundaries in cars, rockets, computer networks, and banking is quite impressive in that context. We can quibble about PayPal, but I am being consistent in ignoring the charging network and self driving system etc as those are emerging fields. Clearly he’s not operating alone, but few people did.
Newton is perhaps the most obvious example of significant progress in independent and established fields and IMO far more impressive, but it’s really not a long list.
The requirement to list someone isn’t to judge that person it’s to judge your standard and demonstrate a flaw in your argument.
Purely objectively, Musk is the second richest person on the planet to the possible exclusion of dictators. That in and of it’s self is rarified air. Winning the silver in say diving isn’t nearly as impressive simply because the is so much less competition. The Marathon is perhaps a closer analogy except that’s a more question of genetic gifts than pure grit.
Doing that and starting a single business like Bill Gates is a true accomplishment, but he’s done multiple independent and highly successful companies. That’s rarified air over the history of humanity.
You personally might not value such things, but that’s simply not an objective standard. Doing that and only making high speed low latency global internet available around the world would be enough on it’s own to be worth talking about. This isn’t a small thing a lot of people failed to tackle the exact same problem. Except Starlink alone isn’t what people think about which is a sign of just how much the guy got done.
The guy literally launched his own car (that his car company produced), into Mars orbit using a rocket that his other company built.
Both of these companies have become the largest companies in their respective fields and can arguably be labeled the two most complex and highest barrier to entry industries. Sending humans into orbit has only been accomplished by large nation states, no company has even done that besides SpaceX. As for car manufacturing, it’s been over 50 years since a new competitor has entered, and only Ford and Tesla have never gone bankrupt.
To anyone trying to argue this isn’t the most accomplished human alive, give me a friggin break
Lots of overlap on the lists, but I decided not to doubly-list individuals -- and the lists are by no means complete, I favored recent centuries.
That said, Musk is tremendously successful and awe-inspiring; this pissing contest is a miserable thing to participate in -- but I still think it's important to point out that Musk isn't the only awe-inspiring success ever.
I'll put it to you another way: if Tesla vanished tomorrow, life wouldn't really change. There may be a few brownouts in a corner of the planet as Tesla batteries vanish and power generators kick in, but life would largely be the same. EV adoption would continue sans Tesla just fine. Self-driving would probably even improve because there wouldn't be the risk of one greedy company, guinea-pigging people and ignoring laws for the sake of share price and potentially bringing the wraith of regulators on the sector.
If Amazon vanishes, Western civilization collapses. That's it. There'd be riots on the street. Critical infrastructure would cease to work for months. Entire supply chains would fall apart and billions of people would be impacted negatively. And it's not like the world would go on. It would take years for competitors to spin up the capacity to replace the void Amazon vanishing would cause in multiple industries.
And the beauty is that while Musk is clearly obsessed with Bezos, I don't think Bezos really thinks about Musk much at all. Just like we can tell that Musk checks the share price multiple times per day, but I suspect Bezos has no idea what Amazon's share price is at unless he's selling to fund other ventures. Also importantly, Bezos has set things up so when he dies, Amazon goes on just fine. If Musk dies, that's the end of Tesla.
Oh the guy who has poured billions of dollars into Blue Origin over 20+ years with little to show for it aside from failed legal attempts / overturned patents to attempt to hinder SpaceX, and a penis shaped sub-orbital rocket?
$100M will also barely make a dent in the creation of a real autonomous humanoid robot. Even $100B might not be enough. That's how hard of a problem this is.
It's a fair criticism that Elon Musk would probably agree with.
He attempts extraordinarily ambitious things that very few other people would even dream of attempting. Sometimes things take longer and/or perform worse than he hoped they would. Sometimes he succeeds beyond anyone's expectations. In any case, he keeps trying to push the world forward, and that's the highest calling of any engineer in my book.
I still can't book a 25 minute SpaceX Earth to Earth flight from LA to NY but I wouldn't be surprised if I could in 10 or 15 years.
> He attempts extraordinarily ambitious things that very few other people would even dream of attempting.
He makes stock plays. Then he lets everyone pick-up the hype for him. This is a very real skill, and he is excellent at it. It is not the same as accomplishing things.
He has shipped 1+ million electric vehicles, transforming the auto industry, which will contribute massively to reducing global emissions, etc.
He has delivered rockets that have resupplied the space station, sent astronauts to space, and will almost certainly lead to building bases on the moon and mars, etc.
A lot more than making "stock plays". The blindness of the haters is astounding. There are very legitimate critiques of Elon Musk, but denying his self-evident accomplishments is beyond delusional.
>he attempts extraordinary ambitious things that very few other people would even dream of attempting.
What are you even talking about? Thousands of people have thought of tackling (1) self driving cars and have made steps towards accomplishing it before musk, (2) reusable rockets, man that has been on the dreams of rocket engineers for decades but we didn't have the computing power to do it, (3) a boring machine... Well we have been doing that for over a century, (4) electric cars, well hate to break it to you but we have been doing that for a century, (5) mars colonization, I feel like everyone born after 69 have been dreaming of how to make that possible; see (case for Mars).
Elon musk isn't some gift from god who comes up with extraordinary projects that no-one else is even dreaming of. He is a wealthy business man, son of wealth, who invests his time and energy into driving top talent in the fields he chooses to attempt to accomplish what he promises vaguely on Twitter. And just like Jobs, he takes all the credit for engineering achievements that could never be accomplished by one man in any reasonable time frame.
Okay, other people dream of attempting these things and Elon Musk actually does attempt them. Is that more clear?
Semantic nitpicking doesn't take anything away from his accomplishments.
If you think Elon Musk or Steve Jobs are just taking credit for other people's work you know absolutely nothing about the history of their companies.
Apple was nearly bankrupt before Steve Jobs took it over, despite having thousands of good engineers working there. It was Jobs' leadership that built into the most valuable company on earth. Other people tried to lead Apple to success and failed miserably. Jobs tried and succeeded beyond what anyone thought was possible. If you think that doesn't matter, or he doesn't deserve massive credit for what he did, you're just confused.
Elon Musk put every penny of his PayPal fortune into SpaceX and Tesla, almost lost it all, all his investors abandoned him, the managers he hired failed him. He was forced to take over leadership of both companies to make them successful, which he did to incredible success.
People like to hate on these guys, to belittle their incredible accomplishments, out of envy, spite, or whatever. The haters fabricate the lie that these guys try to take credit for everything their companies do. Steve Jobs and Elon Musk have both praised their teams and collaborators many, many times. There's no way for them to avoid being the face of their companies, but that is in no way evidence of them trying to take all the credit.
But, the fact is that Jobs and Musk are uniquely deserving of credit for each of their companies' successes. They are the most responsible for their success. That doesn't mean they are solely responsible, a fact they were never confused about, have neither has ever said otherwise.
I'm not someone that thinks these guys are at all perfect people, but if you think they're just lucky, or just anyone could have done what they did, you're just extremely ignorant or lying to yourself to protect your own ego.
This man is living every child’s dream. Imagine being a billionaire and thinking what you could do with all that money? Create a space travel company? Why not. Create a FAST electric car company? Why not. Boring company? Sounds boring but why not.
Some people may not like him for his behavior or marketing tactics but hey, they officially spend $0 on marketing. Glad to see this in my lifetime.
It needs to be compatible with all human spaces and human tools. It'll need to be general purpose to keep unit costs low. This is why the human form is the best.
It's not a menial home robot. It's designed to replace humans in every kind of physical labor.
This reminds me of images from the 1900s depicting the future when there were humanoid robots with lots of arms cleaning lots of dishes at once. Of course, it turns out, we did invent a machine to wash dishes but its just called a dish washing machine and is way better at it than humans with arms.
I can't imagine many tasks where a humanoid will be useful. I'm sure you could say that something like house construction could be done by humanoids since there are lots of steps involved but I'm still skeptical that a bunch of specialized tools won't be better at this.
I see it like computers. First there were special-purpose machines, which evolved into programmable general purpose computers. Probably many people said that there can't be so many computing tasks that programmability is needed. I think we'll find out how these will be used.
I suppose I see the computer as different as it doesn't try to copy anything in nature or existing in tech. ie Google isn't setup like a library, YouTube isn't setup like TV was, etc.
I could see a multi-use robot being used in some instances but I cannot fathom why it would be shaped like a human.
Like I said, there are thousands of tools designed for humans. There are vehicles, ladders, protective suits etc. designed for humans. There are lots of spaces designed for humans in which it needs to move. It makes sense to design the robot so that it is compatible with all these.
And there's nothing wrong with human shape, it would be hard to design a better one.
The vehicle thing is exactly what boggles me about this. Are you saying that the humanoid will be driving cars/forklifts/bullsozers? What exactly do you mean that its important that a robot interacting with a vehicle is humanoid?
Ok. I am wrong in my predictions a lot but this seems way more complicated than just having a computer interface on a bulldozer or whatever - especially given a lot of the cost to operate them is already one of liability
I could be wrong but I dont think it is reasonable that a humanoid would ever track in price with a washing machine. A washing machine only has a few moving parts while a humanoid is going to be more similar to a car in terms of complexity (though less on parts)
There are some machines that fold laundry. They don't seem that impressive now. But I imagine if they haven't already - laundry pick-up/delivery services will roll these out.
The thing is, that this can work alongside humans. say you have 5 people bricklaying and one of those is this humanoid robot. its not a dramatic change from whats currently possible. They are a drop-in replacement.
Think of it as a stepping stone. Like automated cars, Fully automated highways are a ways off where traffic lights are no longer needed, however automated cars that work alongside the rest of the roads are feasible in the near future.
Yes there are better tools, its just a stepping stone to allow change. to force change.
I don't think they are used en masse but I think if we do get to the point of some type of brick laying robots being used it will surely be something like this that doesn't have the physical constraints of two arms with limited mobility + nowhere to store bricks
Something like putting dishes away --- lots of carrying things short distances, changing heights, changing angles, reaching around.
I don't think machines are very close, but bipedal locomotion with human arm like articulation and grasping softish fingers is a reasonable way to get there.
I wouldn't say that dishwasher is _better_ at washing dishes than a human. More quantity? Sure. More quality? Not based on my experience. Lots of stains just don't come out in a dishwasher.
If you can take a restaurant that used to have to have 10 dishwashers and fire 9 of them because a machine works fast - you can keep one guy around to do the tricky stains. Sure is a lot cheaper than buying 5 humanoid dishwashers.
While there are a couple stains a dishwasher has trouble with (typically things that need a presoak, like dough or parmesan cheese), I find that the average quality of a dishwasher wash is way better than doing it by hand, where it's easy to miss a spot, etc. I've also broken several glasses/dishes washing by hand, whereas my dishwasher has never broken anything.
Not even quantity. I can do dishes faster than my dishwasher if you compare end-to-end times. Now, some industrial machines are much faster, but for home use, if I had a humanoid robot that could do dishes, why would I waste space on a dish washing machine?
You have cars (which are basically a box on wheels) smashing in clearly visible objects, and you expect a fully functional walking and handling humanoid robot to fold your laundry or even carry a box from one floor to another. I don't think any of us will still be alive when that will happen.
Really? How many businesses practices are codified using anything other than human-like movement? To expect anything else is the outlier. Any human doing any job in physical labor without specialized design for non-human machinery/robotics is on notice.
>To expect anything else is the outlier. Any human doing any job in physical labor without specialized design for non-human machinery/robotics is on notice.
In fields of high automation, like cars, production lines are almost entirely automated. Where you find areas that humans are needed isn't because our fingers are uniquely adapted to any sort of problem - it's because the problem the human solves would likely need AGI for a robot to do. I don't think anyone is anywhere close to tackling AGI.
I feel like a lot of factories are going to go in this direction. Lots of complex automation but ultimately very simple mechanisms with a strong framework to make sure nothing goes wrong.
Is a cashier/bagger physical labor? Is stocking the shelves of a grocery store? What about a truck driver or unloader?
I would say all these jobs would be automated but 0 of these would be automated by a humanoid. Driver/cashier/baggers will be computers with sensors and unlouder/shelfer will be robots more similar to that vid. Things like construction I think will be assembled ahead of time and brought to side with more automated cranes and sensors.
The number of jobs that seem like "this can be automated but can't be automated by a robot and isn't worth it to build specific robot for but IS expensive enough to use a robot that will always be more expensive than a car" seems small to me.
I would hope that they can use a framework that's general enough to adapt to other form factors. So, making a humanoid is a way of giving themselves a difficult enough initial problem that they're less likely to be fooled by early success: the humanoid robot will have to self-balance, walk, open doors, and do a lot of the complex things that humans do without thinking in order to be considered a minimum-viable product.
In the long run if this pans out, I'd expect to see farming and weeding hexapods, equipment-hauling centaurs, wheeled platforms with arms for getting around office spaces, construction bots with long-fingered hands and asymmetrical multi-jointed appendages, and so on. If the AI has a general capacity for motor-skill learning, then you can just give it whatever body is most optimal for the task at hand.
I guess the difference between me and a billionaire is I don't try and turn all my dumb ideas into products. Eventually if you throw enough crazy ideas out there and refine them I guess some of them work. I am still awestruck at the stuff SpaceX does.
Probably anyone can luck into a bit of success but continued success paradoxically requires a willingness to fail spectacularly.
I think this one can be added to the hyperloop, solar tiles and full self driving but what would I know.
It's amazing, isn't it? Somehow the dancer in the robot unitard wasn't an "emperor has no clothes" moment, or more aptly a "Nikola pushing a truck down a hill" moment.
That's on top of the news that the Cybertruck just got delayed. Again. We're approaching the two year mark of Tesla taking deposits for a product that they don't even have a final design for.
Man, you really lake a sense of humor. I was literally a joke. Comparing that to a scam is the dumbest thing I have read in these comments.
> That's on top of the news that the Cybertruck just got delayed.
A major industrial project got delayed by a few month. Instantly call the police.
And this was the first delay btw.
> We're approaching the two year mark of Tesla taking deposits for a product that they don't even have a final design for.
So what? The money is there, you can instantly have it back. It is quite common practice for many companies both in the car and in other industries.
I literally have books, games, computer hardware on order, some already paid for that will not arrive for a long time. I literally have a kickstarter comic book that is delayed 4 years and yet I don't want my money back.
This product will be production constraint for a while, if you want it early, you have to reserve your spot for 100$, that doesn't seem outrage in the least.
> Somehow the dancer in the robot unitard wasn't an "emperor has no clothes" moment
You can't be serious with this line of thinking.
In case you missed it, here's the key difference between the dancer and the Nikola One being rolled down a hill: The dancer was a silly joke (Elon even said "unlike Dojo, that wasn't real") and Nikola One being rolled down a hill was a phony demonstration. Trevor Milton didn't wrap the video saying, "Haha, just kidding."
You might not resonate with the particular sense of humor that has someone come out and dance, but that's a matter of taste. The fact is, it was intended as a joke.
If you watch the presentation it looks like something they thought of 5 minutes before the presentation. Doubt there's even a prototype. He didn't know what to say about it, just showed the slide and then when people asked questions about it he looked like he was considering everything for the first time.
I have no knowledge of what kind of logic is used in their FSD code, but I don't see how they can derive much of a motion planning algorithm for a humanoid robot from it. Does a car have to deal with self-collision (accidentally hitting yourself in the face with your hand) or singularities (aka "gimbal lock"/"wrist flip")? Does motion planning for a nonholonomic system like a car easily translate into a holonomic system like the human body (maybe it's easier to go in one direction vs the other, like holo->nonholo or vice-versa)?
Is it just me that’s completely creeped out by how it looks? It’s supposed to be humanoid but yet it’s missing all facial features and the black head with completely white body doesn’t make it any better.
What's with all the cynicism and negativity in this thread?
Jeff Bezos launches a rocket - it's not impressive enough of a technological breakthrough for people who watch and judge from the sidelines. Jeff Bezos isn't using his money right.
Elon Musk spends his money taking moonshots and succeeding a lot of the time, announces that he's working on an extremely exciting science-fiction-level technology - it's not realistic enough, he's sure to fail this time.
It's like there's 1% of people who are trying to take advantage of the amazing time we're living in, working hard, often succeeding, making things and getting things done. And everyone else just watches, judges, complains, and tries to score some points by dismissing the achievements of others. Surely if THEY had the intelligence and discipline to spend decades building businesses and make billions of dollars, THEY would have spent all that hypothetical money better.
What does a person have to do to impress you people? What happened to the culture of respecting and encouraging human achievement in America, to optimism about the future?
Personally, it's the overselling of half-baked ideas what I find annoying. I admire Musk for his entrepreneurship but at the same time I don't appreciate being treated as an idiot.
Tesla's announcements leave me with a sinking feeling we won't see anything for at least 5 years after they're announced, if ever. I miss Jobs and his iPod, iPhone, and nano announcements. Those were huge and they were available within months after the presentation.
I loved the no-frills reveal. I’m surprised it wasn’t a collaboration with Boston Dynamics. To paraphrase Elon: “We have the vision and AI so we might as well ship Optimus. Don’t worry, you can outrun it, if you can run faster than 5mph! Coming out sometime next year.” Looking fwd to it.
Full Autonomous driving is not in horizon, but Tesla can repurpose tooling, hardware and some software they designed.
Every time trouble ahead, do something completely new. Overpromise and underdeliver.
No, it will not do shopping for you like he promises in the presentation. Musk had a plan to fully robotize Tesla factories and it failed miserably and set Tesla back on their goals for several years. Making humanoid robot that can work in uncontrolled environment is task even harder.
At this stage electrek is nothing more than a glorified Tesla fan site. Tesla would announce a new genetically engineered, AI-powered cow breed, and Fred Lambert would find a way to have it all make sense.
> Tesla announced that it is actually going to make a humanoid robot, called Tesla Bot, and it will be able to grab your groceries for you and perform other menial tasks.
So, if this anything like Elon Musk's "full self driving" promises, Tesla's actually going to deliver wheeled robot that pushes the cart behind you while you do the grocery shopping, and occasionally knocks over a endcap display.
Why are people so obsessed with this. I mean literally he said he was not gone work on it. He said that literally from the beginning. Yet people think 'He never build Hyperloop' is a great comeback to anything any of his companies announce anything.
He didn't build the thing he said he wouldn't build. Wow, really got him with that one.
I haven’t seen a single substantive comment in this entire thread. Way too much emotion.
Why would he announce this robot if it’s real? If his intention was to turn this into a real product, what purpose is there to announcing it before he’s started?
Even if he were just lying to generate hype, wouldn’t it make more sense to make a prototype and show short clips of its parts moving and actuating? There doesn’t seem to be any scenario where the announcement alone has any justification. Especially when he’s being criticized all the time for these announcements.
Tesla has nothing to do with rockets, so that’s all irrelevant. Tesla does not make the best electric vehicles. I doubt they’re sustainable in the future with more and more competition onboarding. For example, other car manufacturers rev their models every couple of years or so across dozens of models across several geographic regions. Tesla is clearly not able to do that with their existing three models. I guess we’ll see.
> here is still no car that closely challenges 2012 Model S. Tesla has a production problem not a demand problem.
Early gen products are often overbuilt to the point that no future products, from any manufacturer, can ever compete. Early on, it has to work. A decade later and engineer #3451 has eliminated another $0.10/unit by making all the washers 0.05mm thinner which according to the analysis of real world conditions gathered since launch means that it still lasts the 5 years it's supposed to.
The 2012 Model S had infinite free electricity at Superchargers, for instance.
More importantly can it give a good blowjob? Sex is going to be the predominant task of humanoid robots. There's no other reason for them to look human; other form factors would be more efficient for most tasks.
you can look at the latest boston dynamics demos to see what the cutting edge in humanoid robots is. This form factor, with a working prototype, shown in the next year is absurd.
Have you seen the Boston Dynanics videos? If Tesla can do even minor improvement, say with how they do the joints or packaging it better - deciding different constraints to use as guidelines - then it doesn't seem too far fetched; price point will be another factor.
Except Boston Dynamics is a large well-funded, dedicated operation with probably the best robotics people money can buy. Also, they've been it for years and even then their top end humanoid stuff is still super bulky compared to what Elon showed today in their concept.
Also considering Tesla use all 3rd-party assembly-arms in their factories to build their cars, I'm having trouble seeing Tesla shipping anything this sleek and compact to a consumer market..
My guess is he probably figured it'd be cheaper to spend the money on development, especially since it'll be a redevelopment from the ground up following different design-engineering principles as new constraints; including designed to be mass produced which BD robots don't seem streamlined for and are arguably overkill for general use (and not needing to handle heavy loads nor do parkour etc).
Boston Dynamics is one example, but there are loads of examples in manufacturing where the conditions are very tough for people to work on certain types of labor. This is to the point where at times you may have seen references to exoskelatons or specialized equipment to help with repetitive tasks during welding/bolting in a manufacture line that a large robotic arm can't do. A humanoid on the other hand...
The robot clearly not. And FSD yes, but they need to recognize the system will never be level 5. Basically as a company you don't want to be responsible of system failures.
The new frontier these days seems to be machine designed to manipulate human emotions. Oh look, it has an animatronic face, so it's a person! Let's consent to a sybil attack against humanity because we can't distinguish a bunch of plastic from a person.
Tesla 201x -> LiDAR sucks. Humans don't use LiDAR to drive.
Tesla 202x -> Yo guys, look at this vector map we've created with 8 cams and a PS4. But forget about this, you know what ROBOT, LOOK AT THIS ROBOT NOT THE 7K YOU'VE SPENT ON FSD, LOOK AT THIS ROBOT IS FANTASTIC. LOOK HOW THE ROBOT DANCES. ROBOT.
I think it’s an intentional distraction. The vector map stuff is a reply to last week’s news cycle where the FTC and
the NHTSA are both probing the safety of the self-driving systems, but the robot announcement is to make sure no one discusses the “gosh we made the AI better” claims.
This event was planned way before the recent probing.
This was an event to inspire and recruit bright minds.
I doubt it has anything to do with FTC or NHTSA, or even Tesla owners. It was not to announce a new version of the FSD. It was job fair.
I'll pick this up in one of the million+ robotaxis nation wide on my way home from the $800 million shrine we've built in Buffalo, NY that's pumping out solar roofs by the dozen, as they rapidly spread across America. The way the Full Self Driving unit is mounted in the chest is amazing! And the demo unit is so well designed its like a mannequin!
Just looking at one angle here, among other things, this puts other car companies on notice that if they don’t support full self driving, Tesla will supply bots that can do it. Not this year, not next year, maybe not even in five years, but eventually.
That puts the pressure on for the other auto companies to either spend R&D money to compete, or prepare a budget for licensing from someone (not necessarily Tesla) but either way it’s dragging the industry forward.
It also can be a valuable product in its own right which will only get more in demand as human drivers start to be forbidden or heavily taxed on some roads for safety reasons.
That’s not even mentioning the non driving applications. But just for driving there is a lot of legacy hardware that will still be in use for years, and a humanoid form is what they were designed for.
BTW before anyone thinks they will point out some irony, Elon has stated he’s not against the use of petroleum products in many cases. Anyway legacy vehicles are a reality. And they will need safe drivers.
I did say there are other uses. This post just focuses on one use case, I do realize.
SpaceX and Tesla are great achievements, but they are primarily stories of successfully implementing and improving existing technologies with innovative engineering and business practices (like iterative design for rockets), to capture markets that were ripe for disruption because they were controlled by a few large players who had no real incentive to improve.
Problems like level 5 self-driving and humanoid robots, in contrast, are different. We don't see these things because they are hard problems that require basic technological breakthroughs that may not yet exist. Self-driving is one of those things that's relatively easy to get 95% of the way to, but that last 5% is nearly impossible. And useful humanoid robots are probably a generation or two away from being possible; there are just all sorts of fundamental issues like energy storage, object manipulation, locomotion, etc, that are all so far beyond the current state-of-the-art that they're almost science fiction.
I don't see this happening in any useful form in the next 5-10 years.