Tesla CEO Elon Musk unveiled Optimus, the Tesla bot, a few weeks ago at Tesla AI Day. It was a slightly underwhelming event, but showed good progress in just 6 months. Interestingly, Musk said Optimus would be priced at around $20,000.
Of course, he’s been wrong about pricing before. (Where’s my $35,000 Tesla?)
But Musk and Tesla have also achieved significant milestones in product design, manufacturing, and AI for self-driving vehicles. That AI will be a core component, of course, in Tesla Bot as well.
A working, useful, functional humaniform robot has been a dream for a long time. Xioami threw its hat in the ring with CyberOne recently as well. But is Tesla the only company that can pull off humaniform robots with strong AI that can do multiple tasks?
In this TechFirst we chat with Robert Scoble and Irena Cronin from Infinite Retina about what Optimus, the Tesla Bot is, what it can do, and whether it’s possible.
TechFirst podcast: is Tesla the only company that can pull off Optimus, the Tesla bot?
Transcript: Irena Cronin and Robert Scoble on Optimus, the Tesla bot
(This transcript has been lightly edited for length and clarity.)
John Koetsier: What do you think about the Tesla Bot? Is it super cool? Is it kind of lame? Is it our potential salvation or just a scam? Or is it just something that will take maybe a decade to actually deliver? Today, I’m chatting with Robert Scoble and Irena Cronin from the consultancy Infinite Retina. Scoble has been the first to see Siri, drive in a Tesla, experience many other tech products. Irena is CEO of Infinite Retina and has done many other things in technology. Welcome to both of you.
Robert Scoble: Thanks for having us on.
John Koetsier: I am super pumped to have you guys here. Irena, let’s start with you.
Irena Cronin: Oh, wow.
John Koetsier: What did you think when you saw the Tesla Bot Optimus?
Irena Cronin: Well I knew that it was gonna be a raw product, right? It’s not something that is ready to go. It’s not something that’s ready to go in any, you know, measure. They only had six months to work on something. So the actual physical product that was there wasn’t the wow factor really, except, yeah, six months is a short time. I am much more impressed by the AI capabilities that they have and all kinds of things that Tesla already has for it that others don’t.
John Koetsier: Cool. Robert, your thoughts?
Robert Scoble: Yeah, I echo Irena. I mean, why am I excited about the Tesla robot today? It’s the business model that Tesla brings to the robot. There are a ton of Silicon Valley companies that failed at building a robot — from Willow Garage to most recently Giant.AI, which was funded by Bill Gates and Vinod Khosla and already is out of business — building a humanoid style robot for workers, using very similar AI that Tesla’s trying to build. But Giant didn’t have the money to complete the AI and get it to the market, get it into factories and get it manufactured. Tesla does.
Tesla has a business model of selling vehicles and is quite successful at it now. And so, that stream of cash is paying for all the AI researchers we saw at AI Day. And, I think Irena and I talked like the next day after AI Day and we both came away impressed with the effort that Tesla has. The number of people that were on stage, the kinds of AI that they’re doing and the kinds of products they’re building…my Tesla full self-drives now, drives me around town. It’s the most amazing piece of software of their lives, so that gives Elon the cash and the credibility to build the robot.
John Koetsier: Some of that cash must be coming from the $15,000 that full self-driving is now put at but . . . you mentioned software and hardware. There’s two components here, right, and I’ve talked to roboticists and they’ve said, “Hey, you know, there’s a reason why typically robots come in function specific formats. A Roomba is on the ground, you know, a dishwasher, if you want to consider that a robot — it’s probably not, but it — you know, it’s a box, right?
Robert Scoble: At CES I saw window washing robots. That’s all they did. They stuck to glass and they went around and scrubbed the glass. They were made in China.
John Koetsier: Exactly, exactly. Getting the hardware platform for a human-sized, human-shaped robot is so much more challenging. The AI piece, they can borrow on what they’ve built and everything like that. It’s a huge, huge endeavor. Do you think … do you think they can pull it off, Robert?
Robert Scoble: I think there’s only one company who can. Because there’s only one company that has an autonomous car network, a car in my garage.
John Koetsier: Mm-hmm.
Robert Scoble: And nobody else has even gotten close to being on my street, much less in the garage, right?
John Koetsier: Irena, let’s say they do pull it off. Let’s say…
Irena Cronin: Well, let me just answer that question just a little, a little more. So…
John Koetsier: Go for it.
Irena Cronin: Basically, Tesla’s not going for AGI, right? That is, who knows how many years/decades away. So we’re not looking for something that’s really humanoid, per se. We’re looking for something that could do several different tasks. Not just one task, not just two, but perhaps four, and then add to it as time goes on. I think that’s very possible.
John Koetsier: Let’s say it does get pulled off. Let’s say they do do it. Elon delivers it a year or two years, five years after full self-driving ships — that’s my little dig [laughs] — we’ll see when that fully ships, but…
Robert Scoble: I think that’s…it depends how you define full self…if you define full self-driving as a car moving without a human in it, right, that’s four years away.
John Koetsier: Yeah.
Robert Scoble: You were saying five years at the low end and maybe a little more on the higher end? Okay.
John Koetsier: Yeah.
Robert Scoble: Irene and I sort of came to the same conclusion. That’s why we titled our research paper “2028.”
John Koetsier: Could be, could be. So, Irena, if that does happen, how does that change our world?
Irena Cronin: Wow, okay. So when Elon presented, he, a number of times, said that the world’s going to change. It’s going to change economics, it’s going to change the way people live and work. And he was really careful not to go into a lot of specifics there because, I mean, it’s kind of explosive if you start to talk about that and it takes away from the actual technology that he was presenting.
But yeah, you know, if you’re talking about, first of all, the deliveries for pizza and the deliveries for food, that’s already happening now. That’s a very easy task. What’s added onto that is the possibility of the robot entering your home and doing something from that point on. And that’s kind of revolutionary because, like you said, we have the Roombas and we have the little robots. They don’t look like anything that’s human, you know, cats sit on them and they break down and all that kind of stuff. Here’s a robot…
John Koetsier: People upload videos to YouTube.
Irena Cronin: Yeah. I mean, but here’s a robot that’s supposed to serve you and look you in the eye and can do a whole bunch of other things because it has Siri and it has all kinds of other AI capabilities. What will happen, I believe, and even if it’s not Tesla, okay,
but I think this is going to happen anyway, people do not like to do housework. They don’t have time to do housework in this day and age, and they don’t want their house to be dirty, you know, generally people don’t like dirty houses. So, if there was something that could do several different kinds of cleaning, not just a floor and have trouble on the carpet, you know, the window washing thing, dish washing if they don’t have a dishwasher, even placing the dishes in the dishwasher, that kind of thing, it would save people a lot of time.
And let’s say, I’m not saying that they would work instead of doing that … I’m saying that they would be with their families more. They would not be as sick as they are because they have no time to do all this stuff. So that’s one thing that could start to happen. Elon’s alluded to factory work. Obviously, I’m sure that he would love to have these in the factory, so at some point it will replace more factory workers. And then the idea that you could actually have a Tesla Bot Optimus in the truck instead of a person. So, yeah, you have full self-driving but then for monitoring sake and also to load and unload the stuff, you could have…
John Koetsier: Last mile.
Irena Cronin: …more robots, right? You could have like three in the back and they’re deactivated, and they activate when [laughs]… I mean, it’s not science fiction. Yeah.
Robert Scoble: They plug themselves into the plugs in the back of the Cybertruck, right?
Irena Cronin: Yeah, yeah. So it’s going to start to replace menial type jobs that basically no one wants to do. And, yes, what will people do then that need those type of jobs? Well, it’s been like every technology, new technology out there, new opportunities open up that we don’t foresee. So there’s gonna always be something for people, you know? I truly believe that.
John Koetsier: I wish I could believe that. I think I believe that. I think our past shows us that. I do think that AI is qualitatively different, not just quantitatively different, and I think that functional humanoid robots are also qualitatively different, not quantitatively different. But let’s get into that in a little bit, there’s consequences. There’s a ton of consequences, we may have to change how we do our society. Go ahead, Robert.
Robert Scoble: Well, I wanted to follow up on the question you asked Irena, because I come at this very differently. I have an app, or I have an electrical panel that shows me the electricity going through each of my wires. It’s a new SPAN panel, has 26 computers that are watching everything…
John Koetsier: $4,000, if I’m not mistaken.
Robert Scoble: Yeah, it’s fairly expensive, so…
John Koetsier: Yes.
Robert Scoble: But it’s an interesting toy. It’s let me see my house in a new way and I started doing first principles like Tesla, like Elon Musk does. When he goes after a new problem he takes a fresh look at it. And when you start taking a fresh look at the house, you realize the washer and the dryer use a lot of energy … particularly the dryer. And when you start watching how people use their washer and dryer, they’re not filling up the washer and they’re not using it efficiently. Therefore, they’re using too much energy and putting too much carbon in the air from the hot water heater that’s in most western homes.
So you start thinking like that, how does a robot get in there and take over the job of laundry, right? For energy, getting the house closer to sustainable energy to make it more efficient. And the humanoid robot makes a lot of sense there, because some people have really tight spots for their washer and dryer, right?
John Koetsier: Yep.
Robert Scoble: But that was made for a human being to fit in there and put clothes into the washer and dryer, and so now a humanoid robot can get in there and do the job and save the person some energy, right?
John Koetsier: Everything was made for a human being, right? Our cars were made for a human being. Our tools were made for a human being. Everything in our house was made for a human being. That’s why a humanoid robot makes so much sense. It’s orders of magnitude tougher than these purpose-built, specific-task robots, obviously, but it’s way, way more interesting. You bring up an interesting point. You live in California. I don’t know if you have solar panels on your roof, but…
Robert Scoble: I do.
John Koetsier: …kind of a dream of mine, and maybe it’s a fantasy. Irena, laugh at me, it’s all good…
Irena Cronin: I’m not laughing.
John Koetsier: You know, have the house that has the full solar roof, whether that’s Tesla’s solar roof if that ever ships in scale, what the heck? What is going on there?
Robert Scoble: My friend has got one, so I know it’s possible to get ’em now [laughing].
Irena Cronin: [crosstalk] And on top of that, like…
John Koetsier: Yes, exactly, but it’s so hard. Or panels, right?
Irena Cronin: Yeah.
John Koetsier: And then some batteries, charge my car off them.
Robert Scoble: Rich people are getting 16 batteries on their homes so that they can go completely for a month without being on the grid, right, and stuff like that.
John Koetsier: Yeah. And maybe that’s insane, but batteries in homes did save California about a month ago. There were major events going on, there was challenging weather and everything like that, and getting power from probably 300,000, 500,000, maybe a couple million locations, saved the grid in California. It’s important, not just for people who have it, but for people who don’t have it so you don’t have, you know, I guess Texas [laughs], right? So that’s kind of a fantasy that make my own power, maybe do some geothermal as well, right? Charge my car off that and I guess charge…
Robert Scoble: Keep in mind your next car, certainly if you buy a truck like a Rivian, or the electronic version of the Ford F-150 or the Lightning, I think they call it, or a Tesla Cybertruck, all three of those have plugs so they can take electrons out of the truck and put ’em back in your house.
John Koetsier: Mm-hmm. . Mm-hmm. .
Robert Scoble: So that means the battery in the truck can power your home probably for a few days, and that’s if you aren’t very conservative about it how you use electricity, right?
John Koetsier: Yeah.
Robert Scoble: If you keep watching the football games on TV or something, right? If you actually change your behavior and use only your refrigerators, you can go for a week on a Cybertruck or…
Irena Cronin: Yeah, but this could also be with the Optimus robot does the same thing, like what’s going on with the trucks and the cars. Because it’s a robot.
Robert Scoble: Yeah.
John Koetsier: Yeah.
Irena Cronin: Right?
Robert Scoble: By way, if you’re in a hurricane, I have a friend who lives in Florida and has a Tesla power wall. The power wall kept him up and running. His front yard was completely flooded. So, stayed up during the hurricane, but if we had to evacuate from the hurricane zone, we would leave the robot behind to watch the house, right?
Irena Cronin: Take care, yes.
Robert Scoble: And now we can talk to the robot and tell the robot what to do until the house disappears, right?
John Koetsier: Yep. Now there’s a lot of skepticism out there, obviously, because, well, it hasn’t been done before. We’ve seen robots that move better from Boston Dynamics. Mostly they’re tethered, some are battery operated for 20 minutes or half an hour or something like that. And there’s a history, let’s be honest, from Elon Musk of promising electric cars at $35,000, well, didn’t really happen, arguably…or promising full self-driving robotaxis for a long time. That said, Irena, what are the odds that he delivers on this within the next 5, 7, 10 years?
Irena Cronin: I think within the next five years he’ll at least have something definitively ready for his factory. So, they would’ve gone through all the motions and testing for all the many types of things that they would need to do in the factory that would demonstrate saved time and would be greatly more efficient than what they’re doing now.
So, and a lot of the issues that have to do with robots has to do with picking up things. So, robots could pick up the same kind of item. They learn to pick up the same kind of item over and over again with computer vision. But if they have to pick up different shaped items, different colors, all, you know, if it has a variety, it’s extremely difficult. But they’re starting to figure that out now. So, he’ll have that figured out, along with other people that are working on it, and he’ll have it in the factory. And what you need to be able to do that in the home, you need to have that, to be able to have this dexterity with your hands. And also he’ll have the delivery stuff, I’m sure, delivery to your door. Yeah.
Robert Scoble: One way I look at it, if you watch the documentary on the DeepMind team that beat the human at Go, at the game of Go. The AI starts really shitty, really, really rudimentary and doesn’t understand the game very well, doesn’t play the game very well. And over time it just learns. It learns exponentially over time. So, over time it starts getting smarter and smarter and smarter and then it gets really super smart, right? The exponent curve starts going up vertically. And that’s where full self-driving is today. Full self-driving, yeah, I bought it four and a half years ago. It couldn’t change lanes. It couldn’t stop at streetlights. It couldn’t, right, it couldn’t even see streetlights. It didn’t even try to do stuff like that. Today, it’s like driving full time, you know, from here to Santa Cruz. It’s amazing, right?
John Koetsier: Yeah.
Robert Scoble: Same AI. Well, different AI, but AI that has gone along a series of steps of development, right?
John Koetsier: Is that running on the same hardware in your Tesla or did you have to upgrade any hardware at all, like the CPU?
Robert Scoble: Yeah, I got an upgrade. When you buy full self-driving they promise to keep your computer up to date for a few years. And so, yeah, I got…I’m on the same computer that everybody else’s Tesla [inaudible]…
John Koetsier: Okay.
Robert Scoble: Right? ‘Cause I had one of the first 10,000 made at the Fremont factory and so it had an earlier computer, and so they replaced that.
John Koetsier: I mean, it’s interesting and there’s a lot of things that we can look at with the Tesla Bot Optimus, but Tesla obviously is super successful right now, doing incredibly well. The solar side of their business is really…there’s work to be done there in terms of production capability and go-to-market capability and pricing, frankly, ’cause it’s really expensive. The battery side of their business looks like it’s actually doing pretty well as well. Super big batteries. I mean, if this piece of the business hits and this can be shipped, right, that is potentially a bigger business than cars, isn’t it?
Irena Cronin: Mm-hmm
Robert Scoble: Well, keep in mind it comes with the car. This is what Irena and I laid out in our paper. We call it “everything as a service.” Today I can get my barbecue as a service. I can order a weekly barbecue but it costs me $10 a meal to have that delivered ’cause a human has to deliver it. When we go to autonomous cars, the cost of that goes way down, right? And it can bring this humanoid robot to your house which can then come in and do things for you. You know, “Hey robot, can you come in and set up the table? Can you clean my dishes? Or can you wash my clothes?” Right? And it’ll tell you what it can do. At Stanford…
Irena Cronin: It’ll come with a menu that you can pick from.
Robert Scoble: Oh, yeah, a search engine, because Stanford University has a robot and they just released a paper where their humanoid robot does 1,000 things already…tasks. So, my expectation is that Elon’s next AI Day has to beat that. Right? Because you got, and you’re a big corporation, you gotta beat some kids in Stanford [laughter].
John Koetsier: Exactly. Yeah. Your thoughts Irena?
Irena Cronin: I think that several things have to be mastered, but not thousands of things. So, you have to figure out what you’re trying to do with the robot and what its tasks are, and so that it could be done with, you know, 90, let’s say 98% capability, 2% error or whatever. And you work towards that. So you have to be very specific in the goals that you want to have with this robot and then they’ll be met. Yeah.
John Koetsier: You know what’s interesting to me is because they gave a very extensive list actually, of things that they were trying to build into it, and it struck me that there’s a menu of actions and capabilities that if you master that — there’s 40, 50, maybe it’s vastly larger, but it’s low, it’s not millions — if you manage that, you can do almost everything.
Irena Cronin: Yes.
John Koetsier: And so their list included: Moving sideways. Moving backwards. Moving backwards while carrying something. Bending and lifting something up. Lifting something up high, right? And all of those require different capabilities from the robotic brain in terms of balance. If I lift a weight here, how do I balance myself backwards? We as humans, we learn that over, what, 3, 5, 7 years or so as babies, and then it’s instinctual. We don’t even think about it. I can hold something at a distance and I know I have to lean back a little bit so that I don’t fall over with it.
The robot has to be taught all that stuff, but as the AI learns everything, all of a sudden, hey, clean up that table. As long as it understands that activity and knows that not everything goes in the garbage, like that [holding insulated cup] doesn’t go in the garbage [laughs], right? You know, it should be doable.
Robert Scoble: By the way, you said something that it has to be taught that. That is not actually true. The AI today is so fast at learning that it figures out how to walk on its own.
Irena Cronin: It does. Yeah.
John Koetsier: Think of it as a figure of speech…go ahead.
Robert Scoble: Yeah, you know, we have to start changing our mentality with this new AI that is starting to learn pretty fast and being used for things that I had no clue what it would be used for.
John Koetsier: Also in simulated environments. Absolutely. Irena, you’re trying to get a word in here.
Irena Cronin: I was going to say the same thing Robert said, that it doesn’t have to be taught. There doesn’t have to be, you know, millions of things in a data set for it to learn. It could learn something on the fly. So, it’s super great that they’ve come to this junction where they could do that. I also wanted to say that with those 40 or 50 different movements and things that they’re trying to get it to do, what’s important is that they could do several different things together … and that’s what makes a really flexible robot.
Robert Scoble: Yeah.
John Koetsier: Exactly. Not one that does [slow, exaggerated robotic arm motions], you know [laughter].
Robert Scoble: These are fast little motors. I should have brought my DJI over here so you can see the Chinese motors and how fast they are at moving and how quiet they are, right?
John Koetsier: In the future, kids won’t know what the robot dance is. They’ll go, “What are you talking about?” [Laughter as Robert demonstrates quick robotic arm motions] Exactly!
Irena Cronin: Nothing wrong with that. Yeah.
John Koetsier: [Laughing] Yeah.
Irena Cronin: Right.
John Koetsier: Now Elon, with typical Elon optimism, gave a price quote for this thing which I believe was in the $25,000 range.
Irena Cronin: $20,000 and under.
John Koetsier: Frankly, since he’s selling self-driving as software for $15,000, I kinda laughed…you know, I’m assuming it’s gonna be significantly more. What are you guys thinking?
Robert Scoble: How much does it cost us if we went to Shanghai or Shenzhen and go to the computer stores, the electronic store there? They have a six-story store where they sell all sorts of motors, and lights, and sensors, and stuff like that. How much do you think it would actually cost us to build 28 motors that…
John Koetsier: I have zero clue.
Robert Scoble: Probably not that much, right? And so, the real thing when you’re a strategist at a big… I used to be a strategist at Microsoft, you start plotting out what is the last 10 years of pricing on a set of parts, right? And that tells you how soon it’ll be a consumer because you can see the price slope coming down to where it’s a consumer product. And robot technology has been coming down for the last 10 years at a fairly consistent rate. So, you can predict, you know, five years from now, it’s gonna be a lot cheaper than it is today. And a lot more capable. What kind of cameras does does the ‘i’s’ have in five years, right? I mean, today’s leading-edge cameras are 8K on a cell phone, right? What are they in five years? 16K, 32K, 64K, right, somewhere in that range. So, what can that thing see that today’s robot can’t? And what does that enable in terms of new kinds of AI where there’s really fine controls? They can see the world in a very fine mesh that have little voxels.
John Koetsier: What a fascinating concept actually because if you think about today’s robots, typically, we have industrial robots. They’re huge, they’re massive. Often they operate in cages to protect people because they’re not smart, they’re not aware, they don’t know that they could crush somebody if they move this way. They cost hundreds of thousands of dollars. Imagine that the components are the cheap part and the software is the expensive part. Huh. Interesting.
Irena Cronin: Yeah. Let me add to that for a second. So, I think the $20,000 figure that he put out there is probably going to be the basic model without a lot of software. So, it’s going to be the same thing with FSD, you know. It’s like you want the complex model? You gotta pay this. But you got the robot now, it’s up to you how you want to configure it. I think that’s what’s going to happen. Yeah.
Robert Scoble: Not to mention…
John Koetsier: Interesting. You want the model that can take care of your kids, that’s nanny mode, there you go. You want the model that comes with special hands and guards on its legs to do outdoor work and do your garden work, there you go.
Robert Scoble: I know. My brother owns a bar and he wants four of these things. So, right there, that shows the kind of demand, economic demand for this kind of robot, right? A robot that can flip burgers, a robot that can change beer kegs, right, that weigh 40 pounds. Stuff like that. Why is a human doing that kind of task in his bar? That makes no sense to me. Get a robot to do that. But is it a different price for a consumer thing, you know, washing my laundry at home versus a business saying, taking care of people in a bar, right?
John Koetsier: Yeah. Yeah.
Robert Scoble: And I would expect there would be a price difference because we need to tax these things to pay for the jobs that they’re putting out of business, right, and help…
John Koetsier: And that’s Bill Gates’ idea, right? To tax robots and use that for universal basic income. And there, we get to what Irena was talking about. It’s true in the past that whenever we have invented machines, looms, for instance, right? And the Luddites wanted to smash them because their jobs were being taken, guess what? We found additional things to do because we expanded into new needs, new desires, new capabilities, new industries. Maybe that’ll continue to happen. I mean, space is opening up. Elon has another little company there that people may have heard of, right? You know, maybe there’s more things to do, more places to go. I don’t know.
Robert Scoble: Well, this is why Irena and I are excited both about robots and about augmenting reality, because augmenting reality is the other thing to do, right? We could now create much better experiences for people than that flat TV behind me, right?
John Koetsier: So, Irena, you’re putting this in the matrix, huh?
Irena Cronin: It’s the good matrix, you know. We’re already in the matrix.
Robert Scoble: Good matrix.
John Koetsier: [Laughing] You’re intentionally a battery. You want it to be a battery.
Irena Cronin: Yeah. I mean, who knows if we’re in a simulation right now, right? But, like Elon Musk says, but let’s do the best that we can while we’re here. Yeah.
Robert Scoble: We’re at minimum in a simplification because your human eyes can’t see atoms. Right behind my house is IBM’s research center where they let me move a single atom across a piece of copper with a machine.
John Koetsier: Amazing.
Robert Scoble: Something that’s impossible to do as a human being. That, it’s reality. The atomic level is reality, right? That’s actual real. We can’t see that as human beings. So, everything we’re seeing is some sort of abstraction.
Irena Cronin: [inaudible]
Robert Scoble: If we could actually see all the molecules vibrating on the table in front of us, we would…or atoms, we would go insane. So, our minds simplify that, right? [laughs]
Irena Cronin: It’s for our own use, you know?
Robert Scoble: Yeah.
Irena Cronin: Yeah.
John Koetsier: Yes, yes. “Ready Player One” comes to mind, many other things come to mind. The interesting part is, as we get more and more amazing capabilities, whether it’s space, whether it’s humanoid robots, whatever it is, we have so many other issues that we’re dealing with on the macro level. War and economy on the micro level [crosstalk]…
Robert Scoble: Where do you think this stuff comes from?
John Koetsier: Go ahead.
Robert Scoble: Where do you think this stuff comes from? The war machine in Silicon Valley, right? Invents everything and then we use in technology. Our internet came from [crosstalk, inaudible] labs or…
John Koetsier: Yeah. Wasn’t it Boston Dynamics that said that, you know, “None of our robots will be designed for war”?
Irena Cronin: Yeah.
John Koetsier: That’ll be an interesting question for…
Robert Scoble: That kills a whole bunch of potential buyers, but it opens up the consumer market. If you wanna be going after the consumer market, you don’t wanna be seen as a killing robot, right? [laughter]
Irena Cronin: Let’s see what happens with Boston Dynamics because they’ve been on the edge of bankruptcy several times so, you know, They have… [crosstalk]
Robert Scoble: They don’t have the business model. They’re not like a Google, Microsoft, a Face… Meta, an Apple that has things cut…
Irena Cronin: They’re research-oriented, completely. Yeah.
Robert Scoble: And they don’t have probably market fit yet, at least at scale, right? I know Trimble, the construction company bought a dog and walked the dog around construction sites and imaged the construction sites. So, I get they have some customers, but that’s hardly the kind of broad-based customer that Irena and I wrote about in the report.
John Koetsier: I think that they really lack the ability to productize and get that product market fit. I mean, we’ve seen guard robots, right? We’ve seen security robots and they’re on wheels and there they go. Obviously, they’re super limited in where they can go. It needs to be flat, right?
Robert Scoble: Yeah.
John Koetsier: If you take Spot from Boston Dynamics, how can that not be a guard dog? That should be an obvious instant fit right there. It can go lots of different places. If you have enough of them, you can get some of them charging from time to time in their own little charge station so you always have the minimum number.
Robert Scoble: The problem is the cost, right? If that thing was $1,000, I’d buy a couple of them.
Irena Cronin: Yeah, that’s right.
Robert Scoble: But they’re far more expensive because they’re using LiDAR. So, this gets back to the LiDAR versus camera.
John Koetsier: Ohh, the old debate [laughs].
Robert Scoble: The old debate, right? But Tesla is doing full self-driving with just cameras and it works. So, anybody who says you need a LiDAR is ignorant or lying.
Irena Cronin: [crosstalk, inaudible] Dynamic…
John Koetsier: Or have a different opinion.
Robert Scoble: I have a car in my garage, whether it fucking drives with cameras. And so, you can’t tell me the cameras don’t drive because I got proof in my garage. So, the end of debate is there. It’s a lot cheaper, right? They just got rid of the ultrasonic sensors. And Monroe just this morning on Monroe Live said that saves Tesla about $50 per car. They’re making millions of cars a year. That’s a lot of dollars at the end of the year, right?
John Koetsier: Yep.
Robert Scoble: Just on ultrasonic sensors. Yeah.
Irena Cronin: I wanted to say about price. So, the dog costs about 70 grand.
John Koetsier: Okayyy.
Robert Scoble: It’s too much for a security guard, right? You could get a German Shepherd for…a trained German Shepherd for…[inaudible, laughter] Right? I went to a… Anyway, so, the rich people can buy German Shepherds that kill on demand pretty much, and they cost $20,000 after training…
John Koetsier: I think they could get Spot down to either the $20,000 that Elon was talking about…
Robert Scoble: He gives the rich people, right? The billionaires, the hundred millionaires. Those people would buy dogs all day long. But I don’t… You know, there’s a revolution coming to cars and it’s happening literally right now, right? Okay, I think it takes four years to get it completed where everybody says, “Yeah, it’s done.” i.e., the regulation happens and people relax about it. It’s not to that point yet. It’s four years away.
So, when that happens, it completely changes the economies, the world economy, you know, the richer economies at first, right? But I went…Uber was invented right in front of me in a Paris snowstorm and a few years later, I went to South Africa and was meeting with women who were telling me how much it — who lived in the slum there — how much Uber had changed their lives because the taxis wouldn’t come to the slum before Uber showed up. And now Uber is taking them to work and they can earn more and get out of the slum, right? That’s with a human-driven thing. Wait until the effect of the autonomous network of cars is worldwide like this, and it won’t take that many cars, right? An autonomous car can drive 10 people a day around, right?
John Koetsier: Or significantly more. Absolutely.
Robert Scoble: You know, let’s just make it easy to understand, instead of 10 cars, you only need one car, right, for a neighborhood.
John Koetsier: You better make it up on services if you’re a car manufacturer.
Irena Cronin: Can I add… so there’s lots of delivery-type robots right now like Starship and they’re doing really well actually. And they’re ones that, you know, have the larger kind of truck that’s self-delivery. One went under, now there’s a new one that’s being made that’s doing super. You know, this robot, who knows? It could go down the street, maybe on specially-made wheels. And it has its own partition to go down…
John Koetsier: It could use a bike.
Irena Cronin: Right? Yeah. I mean, seriously, the car [inaudible, crosstalk]
Robert Scoble: It can order its own Cybertruck, right? This why having…
Irena Cronin: That’s right.
Robert Scoble: …autonomous cars is so important.
Irena Cronin: Whatever…
Robert Scoble: It has to happen. The first autonomous car has to happen, then the humanoid robot comes together. That is the package that’s gonna really change everything about our lives.
John Koetsier: Go ahead, Irena.
Irena Cronin: No, no. Just to say that once you have a robot that can reason within certain boundaries for your needs, it’ll figure out what’s the best approach to deliver something to you, whether it’s a large thing, then it gets a Cybertruck, whether it’s a small delivery, it could do it on its own. You know, so it makes up its own brain, its own mind, a robot mind, of course, what to do.
John Koetsier: I love that idea. And I’m eagerly anticipating it and I look forward to seeing it. I’m still waiting for frickin Siri that was supposed to be…I was supposed to be able to tell Siri, you know, “I wanna go on a vacation. I wanna go to Hawaii, find me a hotel, find me a rental car, find me a flight, and I wanna pay somewhere around here at this level,” and then just go off and do it. And I haven’t gotten it yet, and that was part of the promise. So, hopefully, all these things will come.
Robert Scoble: New Siri is coming.
Irena Cronin: Augmented reality Siri.
John Koetsier: Oh, I don’t have to go to Hawaii. I could just lay in my bed, get hooked up to a catheter, and [laughing]…
Robert Scoble: [crosstalk, inaudible] I know a little bit about Siri… I had dinner with a guy who runs Siri at Apple eight years ago and he told me…I asked him, “What are you learning about the world of Siri now that you’re at Apple?” And he said, “I’m learning that Google is kicking my ass.” And I’m like, “How do you know that?” And he goes, “Well, we instrumented Google’s AI, Google Assistant, and Google Search, and we instrumented Siri and Google is learning at a faster rate. And anybody who understands AI knows if your AI is learning at a faster rate, it beats everybody else.”
John Koetsier: It’s game over. Game over.
Robert Scoble: And we know Siri sucks, because of that. They knew eight years ago it had to be rewritten [inaudible]…
Irena Cronin: Yeah. But there’s a new Siri coming.
Robert Scoble: What’s that?
John Koetsier: Yeah.
Irena Cronin: There’s a new Siri coming, so you’ll get your wish, John.
John Koetsier: I hope so. I hope so. It’s been a real pleasure. I know you guys have to run. Thanks so much for chatting, and talk to you guys soon.
Irena Cronin: Thanks, John.
TechFirst is about smart matter … drones, AI, robots, and other cutting-edge tech
Made it all the way down here? Wow!
The TechFirst with John Koetsier podcast is about tech that is changing the world, including wearable tech, and innovators who are shaping the future. Guests include former Apple CEO John Scully. The head of Facebook gaming. Amazon’s head of robotics. GitHub’s CTO. Twitter’s chief information security officer, and much more. Scientists inventing smart contact lenses. Startup entrepreneurs. Google executives. Former Microsoft CTO Nathan Myhrvold. And much, much more.