Starship, Tesla Robotaxi, Satellite to Cell and Quantum Computing Breakthroughs

Episode Transcript (computer-generated)

Paul Spain:
Hey, folks. Greetings and welcome along to the New Zealand Tech Podcast. I’m your host, Paul Spain. Great to be back with another episode and very cool to have Keith Patton back in the studio. How are you, Keith?

Keith Patton:
I’m great. Good to be back. Yeah, I’ve been here a couple of times before and, yeah, it’s good to be here.

Paul Spain:
For those that are curious where you fit into this big, wide world of tech in 2024, maybe just a sort of a quick intro about your startup, your new firm.

Keith Patton:
Yeah. So I’ll give a bit of background about myself. Well, I mean, I’m a technologist, sort of minor entrepreneur in the New Zealand tech scene. I built and sold a game called Marker Metro back in the. Spent a few years after that kind of getting versed in the New Zealand SaaS world, working for various sort of large SaaS companies doing cloud architecture. And Vivara is a new startup of mine and co founder Nigel Parker and Manfred Schneider. And essentially what we’re doing is building a platform for personalized wellbeing. So this is an application and a back end platform to make sense of your health data and help you make changes in a sustainable way using an intelligent agent.

Paul Spain:
Fantastic.

Keith Patton:
So, yeah, we’re a few months into that and it’s very exciting.

Paul Spain:
Definitely looking forward to delving into that during the show. Big thank you. Of course, to our show partners, One NZ Spark, HP, 2degrees and Gorrilla technology. We really appreciate their support, not only of the New Zealand Tech Podcast, but of the broadest tech and innovation ecosystems that they get behind here. Well, first up, I mean, there’s so many things going on, but one area that listeners will know, I do get, you know, a bit excited sometimes about some of these new technologies coming through. Starlink was one of those technologies that really fascinated me. Any of the sort of this newer wave of low earth orbit satellites, particularly for things like Internet connectivity, and now it’s coming to connectivity with our smartphones. So we’ve got news through that one.

Paul Spain:
NZ have received the green light to sort of formally test Starlink satellite to mobile service in New Zealand. So across te aroa, they’re going to be able to fill a lot of those gaps. Most of the gaps where there isn’t mobile coverage, which is a reasonable chunk of the country, realistically, because we’ve got a small population, we don’t have towns and cities absolutely everywhere. So there’s a lot of those. Those sort of gaps. Now, most of the time not, you know, there hasn’t been too much of a problem because there isn’t too much, you know, there aren’t too much population out in the. In the. In the back blocks of varying sort of, you know, forests and, you know, mountainous areas and the like.

Paul Spain:
But I think it’s something that we’re very curious about and excited about. So, yeah, to see that, that one NZ have got that. That approval to move forward and to test that, I think is encouraging to see. But one thing that we weren’t sure about was just how quickly they would be able to bring this service online. And we still don’t have the full picture. But what we. I guess one of the areas I was curious about was, well, how is all of this going to. Going to actually play out in terms of the fact that we were seeing maybe some regulatory sort of pushback or competitive pushback, shall we say, in the US market where you’ve got T Mobile who have been partnered with SpaceX and Starlink to deliver that initial service and for it to be then able to spread out around the world with carriers such as one NZ.

Paul Spain:
But of course, if the US being their big and probably primary market, if that wasn’t going ahead, then you could imagine maybe New Zealand would be delayed. But what we’ve just seen in the last few days with the big hurricanes that have hit the US and really created emergency situations, unfortunately, people have passed away due to those. But when there is an emergency of this sort of nature, technology can really come in and help. And so one of the technologies has been a lot of Starlink units getting out there to provide connectivity in varying regions and neighbourhoods. So hundreds if not thousands of Starlink units. But also the Federal Communications Commission has given approval for Starlink to provide actually to turn on that satellite to cellular service in those areas that were hit by the hurricane. So they’ve actually been able to, you know, I believe, turn that service on. And so they talked about there being sort of probably on people’s phones.

Paul Spain:
It would come up as one or two bars. It would show quite clearly that this was a Starlink service with T Mobile. So people will be able to see that they’re connecting to that, but opening it up. So it wasn’t just the T mobile customers, basically anyone that’s kind of stuck somewhere and impacted by the hurricane having that connectivity. So, yeah, an interesting kind of. I guess I. Yeah, positive for testing out the new technology during difficult situation in the US.

Keith Patton:
Yeah, it’s a great engineering feat to be able to. I mean, I kind of see all of the Starlink satellites. We have different jurisdictions, have fiber roll lights, laying things in the ground. And Starlink is essentially our Internet connectivity in space. And then you hit these different sort of jurisdiction, cellular frequency. But I think the partner for customers, this idea in New Zealand that, okay, the tech, I think SMS is first that it just lights that up and a customer doesn’t even need to think about, of course you can buy these Starlink units to connect and have your Internet connectivity in general. But this idea that a customer can just be uplifted by the bridge between a cellular provider and the Starlink backbone, like a chorus provider of the infrastructure, is just really, really powerful, as well as having that direct connectivity. So you have this sort of global capability to provide Internet services.

Keith Patton:
Combined with the costs seem to be dropping as a result of the deployments getting more efficient so they can pass that on. There’s obviously going to be some competitive tension between trying to crystal ball gaze as to who swallows who or whether it’s a sort of keep the cellular providers bridging with Starlink and that there’s going to be the sort of commerce commission type competitive aspects to how it plays out in the future. But it’s unignorable as a means with which more rural areas, poorer people and disasters, there’s some tension and sort of where you get. I think we had issues in Ukraine where you get conflicts and things like that as to what exactly the role is of a private infrastructure provider. So there’s a lot of power in Starlink as a global infrastructure provider for Internet connectivity. And I think the supranational institutions need to grapple with what that means in terms of national security and all those things. But from a pure commercial play, I think for ordinary consumers, particularly those that have waited a long time to get on land or underground connectivity, this is just incredible. And it’s pleasing to see that governments are able to move quickly to help their people in times of urgency and can move through political or regulatory issues with speed to help people.

Paul Spain:
Yeah, look, unless you’re in this sort of situation where you’ve been cut off by the hurricanes, it’s probably pretty hard to fully understand what it’s like in each of those situations. But I can imagine would be people where this type of service has really helped them get, you know, with getting back in touch and, you know, with getting some, some help where they need it, where folks are hurt and so on. You know, there’s a. I mean, we’re just. We’re so used to having access to, you know, amazing communications infrastructure. And of course, we’ve had it here, you know, in New Zealand as well. We’ve had things wiped out. And, yeah, that’s creates some really, really, really big challenges, particularly with our medical type emergencies.

Paul Spain:
Yeah.

Keith Patton:
And I think it’s immune to things that are happening on the ground. I think we do have. I think there’s a sort of. We do have issues now because we’re using so much electronic communication of solar incidents being something that. But apart from that, you know, stuff that happens on the ground or under the ground on earth doesn’t impact our ability to move that, you know, data around when we’ve got stuff in space. So it’s pretty resilient.

Paul Spain:
And, yeah, there’s probably a lot that you could delve into, but we are very much moving into a world where we’ve had a level of reliance on satellites for gps and so on for quite a number of years. And in those odd circumstances where there isn’t good gps capability available for a window of time in a particular location and so on, that can be pretty disruptive. But I think we’re going to very quickly get to this world where we’re just used to that complete sort of ubiquitous.

Keith Patton:
I was going to. That was the word.

Paul Spain:
I was wrong.

Keith Patton:
I was thinking is ubiquity. And then once the technology is there, it becomes seamless. And you’re wondering, you’re looking back, oh, people didn’t have Internet in certain parts of the country because the wire didn’t extend. That may in five to ten years be hopefully a thing of the past.

Paul Spain:
Yeah, sort of tied in with this. And we’ll come to sort of talking about the launch and so on and from SpaceX with Starship during the show. But one of the tweets and bits of information that was shared online that just does sort of dovetail into this is that once they’re able to launch the next generation of, or iteration of Starlink satellites using starship. So once starship is actually in sort of full operation, what we’re hearing is that gigabit connectivity will be possible. So really, anywhere on the planet you’d be able to get that sort of gigabit throughput. Now, I think there are some circumstances today where service providers can pay like mega bucks to Starlink to get, and they need very expensive ground infrastructure and so on to do it. And I guess this aligns with the sort of infrastructure that links Starlink back to, say, to different ground locations like New Zealand. We have those local satellite stations that really aggregate all the comms up and down, but individual providers can pay for, say, a ten gigabit connection if you’ve got.

Paul Spain:
And I can’t remember the figure off the top of my head, I seem to remember $75,000. Was it per gigabit? Maybe so it wasn’t per month sort of thing. So it was.

Keith Patton:
We can do it, but it’s gonna.

Paul Spain:
Cost pretty very serious numbers. So putting that sort of context that a service provider, you know, might pay. So ultimately, millions and millions of dollars to get, say, that, or even 75,000 for a one gigabit. Right. Okay.

Keith Patton:
Do you remember how expensive ISDN used to be? And was the t one links in London? I remember, yeah.

Paul Spain:
All those old Internet links. Right. Slightly thicker wire and a fraction of the performance of anything that we’re used to today. In fact, you know, you just wouldn’t be able to cope with one of those connections if you were to, you know, just trying to watch Netflix or something. Right.

Keith Patton:
But then the possibility seemed endless, you know, pre video.

Paul Spain:
So. Yeah, so one gigabit apparently. Apparently coming. You know, that’s. That’s. That’s part of the future with the future iterations. So I think latency dropped one as well is quite.

Keith Patton:
I don’t know if we’re at the stage, you know, from planes where it’s. I can’t remember the figures, but that was coming down as well. So you got a limitation on speed of light. But, yeah, they’re managing to drop the latency. I’m not sure if you can. If Elon can play Diablo four on Starlink just yet, but, yeah, that’s an important factor, too.

Paul Spain:
Yeah, well, I think the newer iterations will become maybe even closer to Earth and bring down that the latency, which for things like gaming and so on, makes it’s not a big deal for most people. But, yeah, I, you know, more helpful. And obviously, you know, there’s. Yeah, there’s an aspect where it wasn’t very long ago where a satellite Internet connection and trying to do, you know, voice and video communications that. That just didn’t stack up because the delays were so big or it was a pretty poor experience. Whereas now it’s, you know, it’s coming to that. Those sort of levels where it’s just another. Another way of connecting to the Internet.

Paul Spain:
You don’t think too much of it.

Keith Patton:
Yeah, it’s been impressive how quickly they’ve managed to increase the speed and drop the latency.

Paul Spain:
Yeah, I saw news there. RNZ sort of reporting on government’s identity check, facial recognition, apparently being sort of cleared of any racial bias. And we’ve seen these reports in the past that those with fairer skin, easier for the technology and the AI to distinguish features and so on, and those with darker skin, there being a higher risk that they would be recognized as someone who they’re not. But apparently the technology that has, I guess, had some sort of a rubber stamp because been used across 70,000 people already across the last year by the Ministry of Social Development. He’s got that sort of stamp to say this is good. And Tim Waldron, this was a report in RNZ. Tim Waldron, who leads the project at the internal affairs department said this is a service we want to have wide scale adoption. I actually went to high school with Tim, but he’s now based in Wellington.

Paul Spain:
But yeah, so that was quite interesting, talking about wide scale adoption of facial recognition technology, because I think there’s probably a few concerns around how we utilize these types of technology. And I think biometrics, if it’s locked into your phone and the software is sort of safe and secure and it’s never leaving your device, I’ve a reasonable level of sort of peace of mind of that, not 100% because, you know, things that can happen and these things don’t always play out how you expect. But I guess I’m less comfortable in these sorts of things being pushed out more broadly because we’ve just seen so many circumstances where, whether it’s private entities or governments and so on, end up losing control of their own data. And, you know, once you, once you lose your, you know, your face, your fingerprint, these sorts of things, it’s not like, oh, just reset your password, mate, we’ll get you a new one. Yeah, you know, you kind of stuck with these things. So, yeah, I don’t know. I don’t know. How do you feel around these things? And, you know, how we should be progressing.

Keith Patton:
When I read that was a couple of things, one, that it’s great because maybe one, two, three years. We talked a lot about bias in AI and it’s a really good example, I think, of where individual countries with different demographics need to apply a strategy to rid the AI of bias as much as possible for these things. I understand this is within MSD, but I think at a national level we need to be cognizant of this across the board when we’re doing anything in New Zealand with AI that’s to do with recognition or identification. So I think it’s great. I think it was 150 people and it’s really pleasing to see that that’s passed so that we’ve managed to at least solve that challenge with this. I understand the technology that’s being used, is it called live checking? So presumably some form of moving the head around and blinking and changing the head, which is a safeguard. But I don’t have any evidential concerns other than it seems to be. My prior would be that the speed with which artificial intelligence is moving and you see some abilities for single images, portraits, famous pictures to be turned into short videos with heads turning and interpolate.

Keith Patton:
You know, it’s actually gent working out.

Paul Spain:
Yeah.

Keith Patton:
That’s getting so good that I worry about the ability of a techno, even a liveness check to be able to keep up ahead of attempts, too. So it’s a cat and mouse.

Paul Spain:
It is.

Keith Patton:
And so that would be my concern, would be the ability for it to be hacked. But presumably those working with the technology are confident. But I don’t think it’s like you’ve done it and it’s done. It’s one of those things that it’s a constant, like any sort of cybersecurity effort, it’s a constant and quite new battle against perhaps quite novel attempts to circumvent this technology and might require more resource and more focus on that at a national level, but it’s pretty cool otherwise from the risks.

Paul Spain:
Yeah, yeah, yeah. And look, these things, it’s always a balancing act because, you know, I talk about this often, we want to get that maximum leverage and uplift that we can from the technology, but we want to minimize, minimize the risks and, yeah, look, I don’t know what the, what the perfect answers are on these things, but, you know, I guess when you, when you look at, you know, more dystopian type of approaches and you say, look at some of the things going on in China and the control, when you’re using a mix of technologies to try and control a population using by leaning very heavily on technology. Yeah, there’s some concerns about that and I guess there’s been a few folks that have contributed to that concern by making content like black mirror and so on to kind of stir us up a little bit. But there probably is some. Yeah, there’s some reality that there are some risks. Right. It’s not like it’s zero risk. And I don’t know.

Paul Spain:
Yeah. How you kind of draw those, draw those lines necessarily on some of these things and say, no, let’s go for it, it’s going to have a much, much bigger positive impact than a negative impact. Right.

Keith Patton:
Yeah, I mean, I have some sympathy with a friend who’s over in Nashville. He’s in the States at the moment. And we both have the world coin, the Sam Altman. You keep redeeming these currencies and I’ve got these coins, but there’s like orbs. So the kind of star Trekky thing is that in order to identify as a human, you have to, you know, stare into an orb, and there’s someone who has an orb. So he managed, and his wife managed to get themselves verified as human, you know, by looking at this orb.

Paul Spain:
Yeah.

Keith Patton:
Which unlocked some more grant. And this is like Altman world’s coin. It’s like we need to be able to say that we’re human. I don’t know whether world coin itself, but I think with all media and our own identification, there’s going to need to be some foolproof way of doing this as a baseline for anti fraud and identification. Quite where they’re staring into a Norb is like, some of the more early adopters amongst are quite happy to do that. But there’s also, you know, we have banks in New Zealand that are accepting voice as a way to authenticate, to gain access, to spend money on your accounts.

Paul Spain:
Hello. Hello.

Keith Patton:
This just really surprises me that someone hasn’t just gone in and gone, just turn it off. It would be safer to turn it off rather than keep it running. Those services still existed. I agree on that. And I just think things like that are in. Banks don’t move particularly quickly, granted. But to have those things in production right now, in 2024, in October, I think, is. That’s quite scary.

Keith Patton:
I think, you know, so turn off voice authentication for now.

Paul Spain:
Yeah, yeah. Good call. The interesting thing in terms of, you know, those that have used this already, you know, 70,000 people within Ministry of. Of Social Development. So you can imagine, you know, if this was something that’s making it easier for you to, you know, get your health benefit, unemployment benefit, you know, et cetera, and you’re under, you know, you’re under some sort of, you know, pressure. Cause, hey, I need to be able to eat. I don’t know whether that’s sort of the right place to test these technologies. It’s probably an easy way to get access to a large number of, you know, a large number of people.

Paul Spain:
And we’ve talked about supermarkets that are using some of this technology and so on. And in those cases where they’re not keeping, I guess they’re only keeping data on those that have maybe caused a crime or some sort of harm. Within a store. And then if they take your photo as you walk into a store or around a store, they’re dumping that data pretty swiftly. So that brings down the risk levels of what happens with the data, but they are still doing it. And again, that’s where you. How do you make those judgments on what’s appropriate? And if we end up with all supermarkets doing it, then suddenly you don’t have a choice. Right.

Paul Spain:
Because every supermarket has this sort of technology. And of course, there is a positive in it for those that work in those supermarkets and for bringing down crime when people that commit those things get blocked from going into the supermarket. Yeah. That starts making life inconvenient for those bad actors.

Keith Patton:
Yeah. I mean, one of the ways to deal with sort of invasive technique. You mentioned sort of China or social is to be like, to have a degree of openness and transparency about the Met, of why you’re doing it, how you do it. You don’t need to show the code necessarily, but to be, for people to understand from the public sector why it’s being done and how so that they can unpack it simply is really important as opposed to, you know, the more secrecy there is. Not necessarily the code level, but like, the mechanisms and how the data is being used. I think that sort of social contract between people, I think people will be okay to do this with health or with their, as long as they understand how it’s being used and that there are promises back. And unfortunately, we live in a country where most of the time, the social contract is upheld, and we can see that. So accountability, transparency is key.

Keith Patton:
Just give us your face and trust us.

Paul Spain:
And I think at this stage, I mean, the things I hear back are, the general feeling in China is most people are actually, they’re all for the facial recognition and so on. And it’s part of what has helped keep their society quite safe at this point in time. So I think you have to look at it through the lens of. Well, actually, those who are impacted by this technology can be largely happy with it. But, yeah, I don’t know what the percentage is who don’t feel good about it because I guess you don’t want to be leaving people out in the cold and disadvantaged. Yeah. Now onto other topics, there was a rather a well publicized event on Friday where Tesla and Elon Musk unveiled kind of the next look forward on where Tesla are going. The robo taxi, the.

Paul Spain:
And I’m not sure about this, whether you’re gonna laugh at me, the reboven Robovan I think in.

Keith Patton:
Sounds like a type of shampoo or something. Rebouven.

Paul Spain:
Yeah, that was the terminology musk used, and I’m not sure whether that one’s gonna stick. I think Robovan probably makes more sense. And then there are also at their event, a bunch of the Optimus robots wandering around. So quite a few pieces to unpack there. I think maybe we’d look at the robo taxi first, because that seemed to be the big focus of the event. And in fact, the stock price had been moving up a little bit, I think, in anticipation of this robo taxi announcement. The results were maybe not liked as much as some, or weren’t as good as what some expected, because the share price has kind of come back a little bit since. But the big things that sort of stood out to me.

Paul Spain:
Talking about a production in 2026. So hedging bets a little bit. A little bit of time for Tesla to. To work out whether the technology’s right. We also have the reality that Musk will often give dates, and that there’s no guarantee that those dates actually mean much or anything at all. Although there’s probably increasing pressure coming on him to be more accurate as time goes on. Cause the more you put out dates that are overly ambitious, the less and less people trust what you say. He was talking about a running cost, and this is sort of us dollars and us measurements.

Paul Spain:
$20 to which. Yeah, I guess. Could that be 30 to 50 or $0.45 sort of per kilometre in terms of a kind of a running cost, an ability for people to actually buy these robo taxis for around 30,000 us, which, you know, by the time you added in a bit of tax and so on, might be 50 or so. K. New Zealand. Yeah. That have a new, new chip in them. This hardware five chip.

Paul Spain:
Now, our vehicles today, and you’re driving.

Keith Patton:
Is it a model 2017 model S. Okay.

Paul Spain:
Yeah. And that’s got the winter. Has that had the upgrade.

Keith Patton:
I had an upgrade a couple of years ago and then I went in to get a service and they replaced all the cameras for free.

Paul Spain:
Oh, okay.

Keith Patton:
So I got all of the. Whatever. Hd camera upgrades. Obviously haven’t installed new cameras. The modern cars have the inside looking at you for eye detection. Fortunately, I don’t have any of that niggling. So, you know, I can put an orange on the steering wheel and just fall asleep. No, no Joker.

Keith Patton:
But, yeah, no, that’s been kind of good that the hardware.

Paul Spain:
I’ve received those updates, which is completely unique amongst automakers. Right. There’s virtually no other automaker that lets you do much in terms of upgrading the technology in your car. A little bit of software upgrades, but not to that sort of extent. So that’s a seven year old vehicle. Mine’s five. You know, the model three I drive is five years old.

Keith Patton:
Have you had that long? Can’t remember you getting it.

Paul Spain:
So it’s, you know, we’ve got the. Yeah, we’ve got these vehicles. The technology does keep getting better, but this was one of the areas of contention, so they talk. So I think ours would be hardware three. The newest vehicles are starting to come through with hardware four into the New Zealand market. Well, I would pick at the moment. I haven’t actually so closely followed that recently. But I think even the model, why we got late last year is a hardware three out of China.

Paul Spain:
They’re talking about this hardware five and the newer ones, but existing ones being able to act as robo taxis. So vehicles like ours, if we want to get a bit of a return on them, in theory, you could put them on the network and rent them out and they can drive around like basically driverless ubers. But one of the questions that came from the audience that then musk kind of. Yeah. Didn’t give a straight answer to was, was whether those existing vehicles with the hardware three are actually going to. Yeah. Be able to be part of these robo taxi fleets. You heard that comment?

Keith Patton:
I did. I tweeted after that, you know, it was just that when he said that, let’s not get nuanced, you know, the joke. I think he knew, you know, he wasn’t poo pooing. And I was just, you know, mindful of the fact that not everyone will be as patient with the delivery of something that was promised. Well, in my case, nearly eight years ago, yeah. But when I got it, I actually did think. I didn’t think anything that was being said about autonomy coming in 2016. In my head, I had ten years.

Keith Patton:
So I said, if I can get this by, what is it, two years now, I’ll be pretty happy. It would be pretty stink if I had to wait longer than ten years. But the core of this is that autonomy isn’t solved. Autonomy has been promised multiple times, which is fine. That’s what Elon does. It’s been rewritten several times, and we’ve had several. I think that the gains that they’ve had through the implementation of putting generative AI in and removing all the command and control code, just explicitly telling the vehicle what to do. But I think there’s something missing.

Keith Patton:
I think that a lot of the AI, the critiques of the straw man, where it’s like the new generative AI doesn’t do certain things, there is, let’s say, neurosymbolic logic or models or a world model that can’t emerge from the language processing. So I think that there’s possibly something missing with the usage of generative AI or LLMs in general, and that the scaling of that, the big societal shift we’ve got is whether they promise to scale that, whether you will get complete reasoning, understanding, and the ability, and that implies to processing vision as well. I suspect that there is some paradigm missing. And if that’s the case, from an investment point of view, the cost of the robo taxi and the cost per mile and all of the autonomy reveals from Tesla are contingent on the belief that that piece will be solved. Some people believe it’s just scale and money, and some people, myself included, believe that there is a technological breakthrough that is required to complete all those edge cases that you might have experienced some of them, and wonder on your trip to the US and thinking how are they actually going to resolve? Is that just more data, or is there something more fundamental at play? Musk’s choice here to not do the straightforward, let’s call it Tim Cook thing. I just say that as a more conservative approach versus a Steve Jobs approach, which is deliver the compact car, people want a 20k vehicle, get another model y hit and do this at the same time. And I think there’s been some sort of internal similar to Musk putting his foot down with, ironically, the SpaceX, how it came down and was held by these little chopsticks. That was him pushing that.

Keith Patton:
And the entire engineering team were just like, you’re crazy. So, you know, that doesn’t mean that because it worked for SpaceX, it’ll work for autonomy. But I think people were a little bit jaded by previous pronouncements and a very hostile media environment for someone who is getting very involved through the expert just on his own political mission. I love that he’s just tweeting about rockets again, frankly. So I’m not sure. But the upside, if the autonomy gets solved, all these things will happen on the robotaxi and the robovac, all contingent on the ability to fulfill their promise of autonomy. So those dates, 26, I don’t see that in production. Real cars, real people, in one, two years, I’m end of the decade.

Keith Patton:
So I kind of thinking slightly beyond.

Paul Spain:
I think it will happen now. I saw something else there. I’m not sure if you saw the same thing in that. The robo taxi, as it’s being called, it was autonomously driving around the Warner brothers. Warner Brothers lot and so on, autonomously. It seemed quite impressive. They’ve got all these bits and pieces of software.

Keith Patton:
Were the cars autonomous? There were some rumors I’m not going to catch.

Paul Spain:
There. Might have been remotely driven.

Keith Patton:
Remotely, as well as the humanoid.

Paul Spain:
Well, I guess.

Keith Patton:
Need a fact check.

Paul Spain:
Yeah, I guess that’s possible, but I envisaged that. I looked at the robo taxi and I thought, this doesn’t look a million miles off what people were suggesting for the name, which Musk is pood pooed, but what some people were calling the model two. So the mass market TesLa and I didn’t. None of the media coverage I saw mentioned this, but it seemed to me as though that looks like the lower cost version. That could be the kind of not quite budget. But I guess if they get it down to 25,000 us, then, you know, this becomes the mass market electric vehicle from Tesla. That is the next step on from, say, the model three and the Model Y. They were talking about a smaller range.

Paul Spain:
I can imagine. Actually, that would work as well with a lower cost vehicle for the mass market. You basically just have to put the. Put the steering wheel back. But the vehicle didn’t look a million Miles off. Right.

Keith Patton:
No. I mean, I think. I’m not sure whether that’s this sort of tension within Tesla, whether they’re doubling down on autonomy in the roboTaxi, but whether or not there will be an option, at additional cost, to add the accelerator and the pedals and the steering wheel for those who feel more comfortable. So it’ll be interesting to see how they go about launching it. I think they’ve kept their options open, but it’s really a double downing on autonomy with the potential to go back to the robo taxi, either calling it that with a steering wheel or calling it something else or doing both. I mean, I always thought, just do both. They’re big enough to walk and chew gum at the same time.

Paul Spain:
You have two options, although it was just two seats, so that’d be reasonably.

Keith Patton:
Well. The roadster’s two seats as well. But we didn’t hear much about that yesterday, seven years ago. True, true, true, true.

Paul Spain:
Yeah, yeah, yeah, you’re right.

Keith Patton:
A small compact car like Nissan Leaf, you can get a few people.

Paul Spain:
So look. Yeah, really interesting. The Robovan, or Robovin, as musk likes to say. I don’t know, whether he was just absolutely, you know, having a laugh with that one. That was kind of odd. The layout that they showed, if you watch the video, 14 people, but they were saying, you know, could potentially take up to 20 people or goods. You know, we didn’t see a lot on that, but just interesting to see Tesla experimenting with some broader designs. But, you know, more than just, you know, a drawing or, you know, a 3d render of it.

Paul Spain:
You know, the fact that they had this thing and some people seem to pull up and I. And get dropped off in it was interesting. The other things that sort of jumped out, the Optimus robots that were being shown off. Yeah, I think that probably caused a bit of a stir online, because, hold on, these things look really, really good. Actually. They look too good. And then there was admission from might have been Robert Scoble, who was asking, like, oh, are you being controlled by a person? Or, you know, all AI and so on, and it’s like, oh, yeah, I’m getting some help here, or what have you. So, yeah, but even that was interesting that they’ve got all the technology pieces together and they were able to kind of give this, this demo, even if realistically there were people doing just about everything that was done by the robots.

Paul Spain:
Musk, again, sort of talking a somewhat accessible price for this type of technology in the future of $20 to $30,000 longer term. Now, we talked about 2026 being the production, supposed production date for robo taxi. Now, an interesting kind of little gate there, because Musk was talking about autonomous rideshare. And I, you know, your existing teslas turning into robo taxis as being 2025. So that would be a way to basically make sure if you’ve got, you know, a year ahead that you’re launching that. That you basically can hold off ever launching these robo taxis until you’ve actually got the autonomous sort of ride share solved the year before it. And if that doesn’t come in 2025, well, they’re probably not going to start production in 2026. And if it didn’t come in 2026 or 2027, you know, and there’s also.

Keith Patton:
Pending regulatory approval, you know, that there’s a few little get out closes. I think it’s Texas and California, right? They’re going to start in said something about two states as priority because they’re potentially the most well covered.

Paul Spain:
And for those that are curious around this type of technology, if you didn’t catch last week’s episode, I talked about my experiences with the current versions of this technology in the US driving a Tesla with what they call FSD, supervised at the moment, which does the very large majority of driving. But as you’ll hear, there’s a few limitations and issues, so you’ve got to keep your eyes on the road and also a chapter around the Waymo technology from Google Alphabet. Yeah. So I guess there’s a lot of wait and see, really, out of that announcement. But I think visually, it was a pretty impressive type of event and it would have been, I’m sure, fun for those that were there to be able to interact with the optimist robots and to jump in the robo taxi. But, yeah, the shine comes off it a little bit if all of those things are just being sort of remote controlled. Right. It takes an edge off it.

Keith Patton:
I kind of guessed at it when I heard the conversations personally, like, I was okay with the rope. I’m very excited about the robots because I think they’ve made a lot of progress with the mechanical and electrical engineering and that looks like they’ve done. They’ve made leaps and bounds. And I think when you’re putting on a show, they could have maybe said something, but it is a reveal. It is at Warner Brothers. So the software is easier than the hardware. I think this was said last week, so I’m quite bullish on that. And also, robots can be.

Keith Patton:
They’re not out necessarily straight away doing your shopping and mingling with each other. So in the house or around the. I think that’s an easier, hard problem than the full autonomy, ranging around making you money with a taxi. So I think I’m a lot more bullish about the robots and filling in the gaps there with software than I am with the robotaxi. And I think that will be a massive market. I don’t want to put the bins out on a Thursday anymore and, you know, clean toilets and things like. So those. If I can get those sorted.

Keith Patton:
20 grand. Sweet.

Paul Spain:
Done. There we go. Musk. Musk. Has a. Has a. Has one unit sold already into the New Zealand market. I’m guessing your first round is probably going to be more in the 150k range, so you might have to think about it more seriously.

Paul Spain:
Yeah, depends.

Keith Patton:
Need to do the gardening as well?

Paul Spain:
Yeah. You need to stack up a bit of a list of tasks. Does it become the new pool boy?

Keith Patton:
Yeah, I’ll just ask my wife for a list.

Paul Spain:
So, yeah, I mean, when I look at that whole era, I am curious around your point, too. What are the gaps in terms of the technology and. Yeah, how can they be solved. And how long does this take? There’s a part of me that does wonder whether based on experiences, and I saw there’s actually a list online where people are feeding in their data of where they found Tesla’s full self drive, supervised to fail, where they have to interrupt and take over driving. And it was maybe 15, 2020 items. And so, yeah, I’m somewhat curious. Well, yeah, for those items, would you put some sort of manual element in the code or would it slow it down too much to be able to sort of say, hey, when you see a sign that says one way, you cannot drive the other way, or when somebody is turning left and the arrow that points left is red, you don’t drive that way. And these are some of the unknowns of whether you’ve got to go a bit further on in terms of the generations of the hardware to have the performance or exactly how that all fits together.

Keith Patton:
I think the prevailing theory without that add on of some new paradigm is that you just need some more data in the regions, you cover it with data, and you don’t put control code in if they, and do this as much as possible and allow the vision only system to do that. And if you’re going down that road, then you just need more jurisdictions, more data, the edge cases, you know? Okay, let’s get more data of that, because that’s not covered. And again, I don’t know, but it feels like that’s whack a mole. And I’m not sure you’ll get to level five like that. Part of me thinks that there’s going to be need, not necessarily if then do this code, but some sort of recourse to a model. My son’s 18. He’s just passed his test. He can drive anywhere in the world now.

Keith Patton:
He doesn’t need the data from China or France. So there’s something in that one. There’s something elegant about the biological neural network, the brain.

Paul Spain:
But even at times, is it even.

Keith Patton:
If we could get, you know, even if we could get some of the efficiencies and elegance of our mind into these systems, I just. The idea that I’m not sure we can solve it with more data. And if we had to go back to if, then that’s where Tesla were five years ago, and that would be an admission of the incompleteness of the generative revision only approach through pure LLM style processing.

Paul Spain:
Yeah, I mean, I guess my wondering was if there was some sort of a hybrid between where they were and where they are now. So they’re mostly leaning in on this training model. But, yeah, we’ll see how that plays out. A couple of other things before we sort of dive into here a little bit about Vivara. Chinese scientists apparently using quantum computers to now hack encryption. I guess this has always sort of been one of the things we’re warned about, hey, quantum computing is coming, and look, your current encryption mechanisms that we rely on for all sorts of things, whether it’s our banking or cryptocurrencies and so on, is potentially at risk. So it was just interesting to see that headline. Often these headlines, when you delve a little bit deeper, it’s not exactly the highest grades of encryption and so on, that they’re hacking at this point in time, but it is, I guess, reminder that quantum computing technology, like every other technology area, is getting ongoing investment and is improving and getting better.

Paul Spain:
So, yeah, we may need to make some changes on the encryption side. When we were chatting before the show, you were talking around that there’s some technology that’s coming through that’s designed to be kind of quantum proof.

Keith Patton:
Well, I mean, often with these things, I always, you look at the media article, and then it’s quite easy to get caught up in some moral panic about the technology being dark or light. Star wars sort of morality kicks in. So my initial thing was like, what are we doing about this? What’s the response? So I saw this NIST North American Institute for Standards and Technology has post quantum, they’re working on enhancing with different style algorithms to make quantum proof new approaches to encryption, which will require us to upgrade military, financial, bitcoin, all of these things that’s going to take years to roll out. But they are aware of and working on this. There’s nothing, as a far as I’m aware, is commercially available to implement. Fortunately, that article and the conditions to establish a quantum computer and the environment is very fragile. So it’s a wake up call, but it had to be performed in laboratory conditions. But it is a theoretical proof that these are current in flight production algorithms are at threat, theoretically, in the future.

Keith Patton:
So the thing, again, like the facial recognition, the thing that doesn’t worry me, but the balance is how quickly can AI, black hat or people who are wanting to do something not nefarious, how quickly can AI plus this quantum computing, not everybody has access to a quantum computer. Costs are high, environment is fragile. But how quickly as we’re seeing in other areas of artificial, how quickly can that transcend? Because we have a limitation on how quickly the world can roll out improvements and the priorities. So that is going to be interesting to see how quickly that catches up. But for now, it’s more of an early wake up call. And I was pleased to see we actually do, us government working on trying to establish some standards around this.

Paul Spain:
I did find some things saying NIST have now released this in recent weeks, their first three post quantum encryption standards. So they’re saying, hey, we can actually tick that box.

Keith Patton:
Now we just need to go and upgrade our servers.

Paul Spain:
Yeah, well, now they’re saying. They’re encouraging the commentary and what was up on their website as NIST is encouraging computer system administrators to begin transitioning to the new standards as soon as possible. I mean, yeah, I would tend to think we’re going to rely on these new encryption standards being built into cloud providers. The technologies that we use, whether it’s open source software or whatever, is provided by cloud.

Keith Patton:
And the things that Amazon and Microsoft are very, Amazon was at the AWS, they’re working on quantum computing as well. So you would imagine the security elements will be feeding through from their research on that into the things that we use within cloud. So to make it as easy as.

Paul Spain:
Possible, now, there was the starship launch also over the weekend. So SpaceX number five launch of the starship. And when you look at the starship sitting on top of the super heavy booster, I mean, you’re looking at something that’s like a skyscraper, right? It was incredible to see, and to see this thing actually even just be able to get off the ground is phenomenal. I guess the biggest rocket ever built by a long shot. And so, yeah, just absolutely fascinating to see that. But really the excitement came with the chopsticks. Tell us why this has kind of really caught your attention.

Keith Patton:
Oh, you know, I know some people who just follow everything that space, I mean, I’m kind of one of these fair weather SpaceX. I mean, I love it. So I came across, I just watched it open mouthed and sort of that feeling of CGI and is this really. But also really proud. Like, just proud. I don’t know. I felt like tweeting thank you to SpaceX. You know, the whole, just because it feels like some, it’s like a landing on the moon moment.

Keith Patton:
It’s not a planet, but there’s something, there’s something incredibly, you know, in the times that we’re living to see such ambition and fulfillment and a bit of luck, you know, I don’t think any of it was, but the idea, I watched a video from a year ago where Musk is standing at the bottom explaining the, you know, lose the bottom, bring the chopstick, we’ll reuse the thing, we’ll save the weight. And, you know, people are thinking that he’s insane and he probably is a little bit insane, but the fact that the sort of sheer compulsion to take the risk, to believe, to get the margins is, you know, it’s very inspiring. And, yeah, the chopsticks, you know, is this. It’s like what a kid would do looking at a small. You just, well, we’ll just bring these, these planks of metal in and then when the thing comes in, there’s a couple of little latches on the side, tiny little bit, which are load bearing and like hundreds of tons. And then they’ll just neatly, you know, just plonk down and do that.

Paul Spain:
Catch the flying skyscraper effectively.

Keith Patton:
And so, yeah, it’s sort of incredulous. Like, I don’t feel like technical person, but I don’t feel the capability that I can sort of analyze what happened. I was in awe like everybody else and then just think I had a real sense of, like, proud to be a human, you know, it’s like, great. A very uniting thing.

Paul Spain:
Yeah.

Keith Patton:
And I think we need a bit of that, you know, with axed the way it is.

Paul Spain:
Yeah, it was really, really, really impressive and, yeah. To just, you know, I think we are watching this development in front of our eyes of, you know, what they’ve been doing with starship and obviously the super heavy booster sort of being part of that whole program and talking about the ultimate goal, something like 200 tons that they would be able to launch into orbitz that, you know, if they had, you know, or when they have human passengers, you know, talking about, oh, yeah, we could put 100 people in there. It’s like what, you know, we’ve never heard of, you know, probably even a dozen people. Even, you know, a dozen people being talked about as a, like, yeah, yeah, the spaceship. No worries. We’ll be able to take a dozen people up and so on. So, yeah, this idea of something that, you know, ultimately would be able to go to the moon and go to Mars and take up these huge payloads is absolutely fascinating. Yeah, totally.

Keith Patton:
It’s create a bit of cognitive dissonance. You know, when you see the robotaxi and then you see that and you say, Willy, won’t I can they? You know, it’s like, sometimes it really works. Sometimes it might not work, you know, but I think people like to paint a brush and make it a. Make a sort of judgment about the whole thing, whether it’s Musk or SpaceX or Tesla, and it’s like, it’s a lot more nuanced, you know?

Paul Spain:
Yeah.

Keith Patton:
But it does give you hope that the stuff that you might consider close to the impossible is actually achievable.

Paul Spain:
Yeah. So what’s the impossible, you know, that you and the team at Vivara are working on? I mean, it’s always exciting to hear about startups and new companies launching, using new and interesting technology, and even more interesting when it’s folks that I know that are involved. So, yeah, really, really keen to hear what you’re doing and what the plans are for the future.

Keith Patton:
Yeah, well, Vivara, you know, on our very small website, says personalized wellness guidance for everyone. To translate that into a couple of problems that people might be faced with. I think there’s kind of an emerging area. I’m sitting here with, like, an aura ring. I’ve got a whoop bandaid. I do have an Apple Watch, but not on me. So this idea of people getting more aware and needing to be more aware about their own health and at lower and lower cost, but getting kind of overwhelmed by blood tests and data coming in from their Apple Watch, what do I actually do with this? And some of the individual apps do help you in their ecosystem. So Vivara is one of the key things is to range across a wide range of data, from your blood tests to whoops and auras and garments, and get this data in like a magic vacuum cleaner, and then range over it and use an intelligent agent to offer a translation service over that data so that you can understand your own health and wellbeing better.

Keith Patton:
We’re not diagnosing illnesses. We’re not a doctor. We want to use the data to inform a person more about their wellbeing. So that’s one side, making it less complex and reducing that overwhelming sense of having disparate sources of data. The other thing is, people struggle with change. Everybody’s different. You open one of these apps and some of them are very more passive, more about social connection, and some are more drop and give me ten, sort of full metal jacket, do the assault course. And people who open those, depending on their personality type, will either recoil or really enjoy it.

Keith Patton:
So for VAR, it’s really important, I think, with the advent of sort of agent based intelligence, that it works really well for empathy and compassion and understanding people. And then once that we have a Vivara guide, so that’s an agent that is going to get to know you and then will be more appropriate in the way that it talks to you to provide insights and recommendations so people will change if they’re given the right type of support for them. I like graphs and numbers in the drop and give me ten. Other people want social connection and conversation and passive reflection. Working around who the person is helps us to be able to make people create sustainable change for them in incremental ways. And I’ve started the company with co founder Nigel, ex Microsoft guy, and Manfred Schneider, our investor buddy. And we’ve all got personal stories that link into this. So it’s when Vivara is kind of transforming, or at least my approach to.

Keith Patton:
I’ve always been in the gym and my father died in 2022, and I put on a bunch of weight and I was just like, I don’t need to go back to the gym and lift some weights. So when I started at Vivara, I was realizing that there was multiple things, sleep, activity, mental health, all these things. And I’d always taken the approach of, okay, I’m going to eat less, I’m going to try and move more, not necessarily both at the same time. And I wouldn’t really think about my sleep, I wouldn’t really think about stress so much because I need to work hard to get money. And then with Rivara, when we were bringing this data together, I realized if I just pull one of, do something small in each of the levers, fix a little bit of sleep, get off Twitter before bed, stop eating the cookies in the evening, sleep’s better, productivity’s a bit up. And then I was better at the gym and I was losing weight more easily. So for me it was an awareness that there’s what levers, you’ve got to change yourself, there’s more of them than you think. And you don’t necessarily need to do twelve week transformations or twelve day detoxes.

Keith Patton:
You know, these sort of rapid, must change, like sustainable change is hard. So the way to approach this is like small steps till they become like brushing your teeth, you know, I’m not going to brush my teeth today, it’s too inconvenient. It’s just something that you do. So there’s various theories around this bit of behavioral psychology getting put into it, and I’m pleased that I haven’t. Did I use the word AI? So I’ve got agent intelligence. So I don’t want to just say AI over again. That’s important. The technology that is there is powerful, yet somewhat unconstrained, linking it with deterministic systems like your health data, understanding who you are fomenting change.

Keith Patton:
The AI needs help with that, so we’re building a platform for the future. And that means that whilst the technology is changing, we need to have a very, we need to be nimble. So some of the challenges are like the technology is moving so quickly, so we have to have our Lego bricks organized in such a way, if the blue piece changes to a green piece and gets three mores, we have to be able to plug in a new thing. So that’s a challenge. But it’s also really good for the architecture of the platform because you’re never assuming that something will be there for two years. Well, some things, but most things, you’re trying to think about how they’ll change. Really healthy. But on the flip side, integrating the technology with using, with our own wellbeing, got a gym across the road and constantly, you know, integrating the measurements, educating our trainers and other people, it’s been a very inspiring and healthy journey for me, as well as doing work in a startup.

Keith Patton:
So that’s been quite transformative. But I think for Varvara, for everyone will be different. The application will essentially cater itself to what you need to help people change.

Paul Spain:
Yeah, I really like that aspect because so often technology is kind of sliced in one way or another and it doesn’t necessarily fit for everybody. So, yeah, if you can have that more sort of personalized. Personalized, that helps you kind of move along in the right direction. But what’s right for, you know, for you, what’s right for me, those things are going to be different. And, you know, as you say, you know, you might like something that, hey, drop and do ten push ups or something. Yeah, that’s not necessarily going to be a, be a fit for everyone. But, yeah, I know from, I mean, just from social feeds between you and you and Nigel, that both of you are pretty committed to your fitness over a fairly, fairly long period. Right.

Keith Patton:
Which is actually a problem because Nigel and I are unfortunately not representative of the wider community. We’re kind of more like confront. We’ll confront things, we will be a bit more driven and that’s not so we have to be mindful when we’re working with our customer research and journey planning that we’re building this for other people, building it for ourselves a little bit, but we can’t assume. So, personalisation, you know, we’ve been using that word for ages in technology and at the moment that means it’s your data. Like, that’s your beats per minute of your heart. But for me, personalisation in an age of, let’s call it agent type intelligent means that the relationship you have with the agent in the application is personalized to you from an empathy, compassion, understanding level, as well as the data itself. And I think personalisation, that is going to transform the experience that you have with an application. And we’re seeing some of these things emerge with, you have chatbots.

Keith Patton:
If you look at cloud or these canvases that are being created to allow smaller sprinklings of AI or agentic intelligence, user experiences are changing, they’re becoming more dynamic. And that’s very, very exciting to me. I have a background in user experience, so that side of it is as interesting and exciting as the leading with opportunity as the sort of the platform or the AI model elements. So the user experience side is key because you see a lot of apps will talk to the AI and then you get a little chatbot on the side, and that’s kind of everyone’s experimenting with it. But when you get a chance to design it from scratch, the experience, it’s invigorating because you can get the agent integration just right. Yeah.

Paul Spain:
Now, understand you earlier on in your journey. Not sure if you’ve raised any external funds yet, but are there any timelines or anything else you can sort of can share for those that are thinking, oh, this is something we’d be very interested in, whether it’s from a usage perspective or an investment perspective and so on.

Keith Patton:
Yeah, I mean, if anyone’s interested, you can go to Vivara Health and you can sign up on the page. We’re going to be calling it MVP or a private beta release, end of Q one next year. So we’ll run that for two or three months, sort of probably April to June, and then we’ll be looking to extend for another 18 months to two years. We’ll be delivered, we’ll be launching sometime next year, as in terms of the first, hey, anybody can grab this. And next year will be all about this year we’re digital health Summit, AI Summit, and we’ll be probably launching with the stall somewhere. And there’s places in Melbourne, Auckland will be releasing and getting more visibility in 2025. But now we’re just scaling the team slowly, using AI as much as possible, so we don’t need to hire too quickly and doing a lot of validation. So these, you know, these problems, these are hypotheses from us.

Keith Patton:
So we’re working to validate them with end users, people that aren’t like me and Nigel.

Paul Spain:
That’s good, that’s good. Excellent.

Keith Patton:
Lots coming out.

Paul Spain:
Well, all the best on the journey, Keith.

Keith Patton:
Yeah, thanks, Paul.

Paul Spain:
That’s exciting. Well, thanks, everyone, for joining us on the New Zealand Tech Podcast this weekend. Of course, a big thank you to our show partners One NZ, Spark HP, 2degrees and Gorilla Technology. And, yeah, thanks again, Keith Patton, for joining us and all the best with what’s next on the Vivara journey.

Keith Patton:
Thank you.

Paul Spain:
Yeah, cheers. Cheers.