Starlink Outages, Government Apps, and Ethical AI

Paul Spain:
Hey folks, greetings and welcome along to the New Zealand Tech Podcast. I’m your host Paul Spain and great to have Brett Roberts joining us again. How are you Brett?

Brett Roberts:
I’m very good and thank you for having me again. It’s good to see you.

Paul Spain:
Always a privilege to chat. Maybe you can remind listeners where you fit into this big wide world of business and technology in New Zealand.

Brett Roberts:
I guess I’ve been in the tech space New Zealand for some decades, quite a few decades I guess and have a, an, an ongoing and huge interest in technology. Mainly about the train wreck about where tech people and business kind of intersected. A lot of keynote presentations and stuff around that. And these days I co own and manage a CNC manufacturing business in East Tamaki which is where I’m coming to you live from at this moment. So there may be noises banging, forklift reversings and yeah. People touring and frying. So yeah.

Paul Spain:
Yep, yep, yep. So the, the, the, the, the CNC machines sort of you know, creating signage and 3D all, all manner of, of, yeah, lots of stuff.

Brett Roberts:
So we’ve, what have we been doing lately? We’ve been machining some very, very large 150 kilogram pieces of timber for the new Fisher and Paykel head office in Penrose just down the road from Fletchers machine. Hundreds and hundreds of those. We do large production runs for some well known companies. We do all sorts of fascinating jigs and molds for aerospace companies that I can’t mention. We do a lot of signage work. We do a lot of 2D work. You know just cutting stuff out of sheets of mdf, plywood. Sure.

Brett Roberts:
Aluminum composite material and things like that. And then we glue things together and make them into 3D. So yeah, we do lots of stuff but three large machines, it’s an interesting business.

Paul Spain:
Yeah.

Brett Roberts:
And interesting space at the moment. The manufacturing space, it’s pretty, pretty quiet, it’s pretty challenged. So you were just kind of taking each month as cubs really.

Paul Spain:
All right. Oh well thanks for taking the time out to join us. Also a big thank you to our show partners to One NZ, 2degrees, Spark, HP, Workday and Gorilla Technology. Well let’s jump in to the New Zealand news side and then we’ll get into the international. Some pretty interesting things going on. The first one that really caught my attention and it was the press in Christchurch that where I saw this is We’ve got company in the south island are so sweet and they’ve gone down the track of acquiring some technology that gets rid of weeds with lasers which you Know, anytime lasers are doing kind of interesting things, I think it attracts some attention from me. I don’t know whether it’s back to Star wars and lightsabers or, or what it is, but yeah, pretty cool use case, I thought, Brett, and looking at the video of the equipment that they’re using, the Carbon Robotics G2 200 laser weeder, which, you know, gets dragged behind a tractor that looks like some pretty serious kit. I’d quite like to have it on my lawn to be able to get the perfect cricket pitch together for, you know, for playing cricket in the backyard.

Paul Spain:
But it might, might not quite fit. But I guess when, when you, you put it into, you know, a context of helping grow, you know, perfect crops and food for New Zealand, that’s. It’s a pretty good use case.

Brett Roberts:
Yeah, that’s pretty cool, eh? I mean, I know there’s other companies out there that have been doing this thing with lasers and weeds, but I think that combination of lasers, artificial intelligence to spot the weeds and then dragging behind a ginormous John Deere tractor is quite a interesting use case. Also a very, very cool laser pointer if you want to, you know, entertain a lot of cats or a really big cat or. But yeah, quite a neat way of doing things and getting, and obviously getting rid of herbicides, which can’t be a bad thing either. So, you know, cheap labor, I think see farming sector more and more things like this. You know, if you think about it, that’s such a great use case combination. You know, it’s. I would have mentioned these days relatively easy to train AI systems on recognizing weeds versus good parts. Lasers are incredibly cheap to produce these days and the auto and you know, autonomous vehicles, not even in dragging behind a tractor that’s sitting on autonomous vehicle, you know, something like one of the vehicles from robotics plus something like that sort of makes a lot of sense.

Brett Roberts:
We’re seeing more and more of those things. I wouldn’t put my hand under it to clear anything out just to be on the safe side, but, you know.

Paul Spain:
Well, hopefully the cameras would recognise your hand and you’d be safe. You were telling me earlier about a somewhat similar scenario where the AI recognition through cameras is keeping, keeping people safe with, with a, with a big band saw.

Brett Roberts:
Yeah, there’s a. There’s a company just down the road from us here in East Tamaki. I think it’s called Kmon K A M O M or something like that. What they do is manufacture absolutely beautiful bandsaws for the meat processing market. Very large things, you know, they’re sort of two, two meters tall, lots of stainless steel in them and the way they work is they have a camera system mounted above where the saw blade is and the operators wear special gloves and there’s an AI system built into it that recognizes and tracks where people’s hands are. And the minute any, anybody’s digits get within danger, dangerous area of the bandsaw, it stops the bandsaw immediately. Quite incredible company that I’d never heard of before and sent to have a conversation with one of people from there late last year and they gave me a little bit of a tour around. It’s absolutely very super duper impressive.

Brett Roberts:
I was blown away with what they built and machining. They do all of their own machining. Got these beautiful machines for machining stainless steel for making the flywheels and all the other bits and pieces just. And a great export store. I think most of this stuff goes up to the usa. Yeah. Another good use of kind of old technology bandsaws and new technology. AI.

Paul Spain:
Yeah. It’s fantastic.

Brett Roberts:
Machine vision.

Paul Spain:
Yeah. Yep. Well, I think there are probably still a lot more use cases that have yet to be sort of fully developed out and some of them will be these sort of smaller edge cases. But when you stack it up and you look at what the benefit is of people not losing, you know, individual fingers or, or more.

Brett Roberts:
Right.

Paul Spain:
That is, that’s, that’s huge. And you know, I think any of these areas where there are, there are health and safety challenges there probably a bunch of them that are, that are. Right.

Brett Roberts:
Machine vision and health and safety go together really, really well. There already are lots of solutions out there.

Paul Spain:
Yeah.

Brett Roberts:
For that sort of thing but you know, to track people and vehicles and keeping them apart from each other and stop trucks pegging over people on the courts and things like that.

Paul Spain:
Yep. Now we’ve had news through that the Department of Internal affairs has appointed two guess sort of tech and tech related firms in New Zealand to build out government application that’s sort of being, being, you know, talked about as a, as a core government app and that they’re going to, yeah. Work together on producing this app. One is Dave Clark, Dave Clark Design. I know them as I see in the media release, it’s called it Dave Clark nz. So you know, they’ve been around for, you know, for 20, 20 odd years. Pretty talented, you know, team there and they’ve grown out to a reasonable sort of size team. I think maybe they were in the direction of sort of 80 people across a number of offices now.

Paul Spain:
I think I Remember visiting them must have been in their early days about 20 years ago. And then they’re working with MATTR, which my understanding is they sit under Spark. I think they’re maybe 100% owned by Spark if I remember correctly. And MATTR have been, I guess building up a skill set around digital identity and digital wallet type tech. And so I guess Dave Clark will create the sort of the end result but they will be leveraging technology from matter. M A T T R to deliver this new government app. What are your thoughts? Does it get you excited that this is coming? Is it going to be helpful to have some sort of centralised government app for New Zealand or does it, you know, I guess there’s a few ways this can play out, right?

Brett Roberts:
Yeah, look, I think anything that helps connect citizens to government and vice versa, centralize, you know, access to those services, make it easier, provide digital identity and lots of things. I think there has to be a good thing. I can’t, can’t see a downside in that. You know, the devil’s always in the detail and getting a disparate number of government agencies to figure out how to join all those things together in a coherent way and delivered environment is always interesting. I’m actually going through an experience at the moment. It’s quite. You’ve just reminded me I have a real me identity as do many of other Kiwis which is coming up for expiring. In fact I think it expires in the next day or two and they have quite a great system where you basically can just take a selfie of yourself and then you take a self video of yourself doing some things and that’s how they reprove your identity.

Brett Roberts:
Well, for some reason that no one understands the no longer works on an Apple phone. Works okay on Android but it doesn’t work on Apple. It’s been down for weeks and weeks.

Paul Spain:
Oh dear.

Brett Roberts:
And so I can’t renew my, my identity. I don’t think I’ve changed but I need to prove to them that I haven’t. And so some of those things I, you know, I think there’s, as I said, the devil’s always in the details. But yeah, look, I admire the DIA for trying to drive this forward. I think we’re behind the eight ball. I think we’re. There’s other countries out there that are way ahead of us with some of these things. But it’s good to see someone taking some initiative.

Brett Roberts:
It’s good to see some folks appointed to drive it forward. But I think some solid quick early Wins would be a good thing and not have this be something that three years down the track we’re still talking about.

Paul Spain:
Yeah, look, I think making things a little bit smoother and easier, as always, always of interest, actually delivering on that can be easier said than done. One of the things that, as we move into a world of digital identity that I do keep hearing sort of different bits of commentary on, there being some risks associated with that. When I look back, one of the things that we’re very used to in New Zealand is, hey, the government has a lot of information about us, but generally it’s been siloed up. So one government department talk to another and you know, we don’t end up with this kind of scary Big Brother type scenario. But there is now this, what feels like an inevitability of actually all of the dots getting joined up and, you know, governments being able to link together all sorts of different bits of data. And, you know, I think there’s a bunch of areas where that’s probably pretty, you know, pretty, pretty helpful. You know, someone’s flying overseas and, and they’re off overseas for five years, but they’re claiming some sort of government funds that they shouldn’t be and so on. But I think there’s also a level of getting the right balance here so that we don’t kind of create a digital net that everyone’s getting caught up in on an ongoing basis.

Paul Spain:
And we don’t want too much data sitting around because we know whether it’s a government or a private entity, you know, no organization’s perfect when it comes to kind of keeping and looking after people’s private data.

Brett Roberts:
So yeah, yeah, look, it’s a really tricky scenario, right? You know, you know, need the data to do things with, you need legislation to protect that data and the people that, you know, that data is about. You know, anything could happen in the future with regard to governments. I mean, you know, who would have thought we’d have a crazed orange redneck totalitarian running the United States, right? In, in some equally crazed ketamine addicted billionaire wandering through government departments, taking all the data from, you know, who knew? So, so, you know, how do we protect against that? And also, I think that aggregation of data sets is a really important point here. Right. You know, when I do presentations and I talk about privacy and data, you know, that there is nothing stopping me setting up a whole lot of automatic number plate recognition cameras all around the country and tracking all of the cars that drive past them. And there are data sets that I could Buy off the likes of spac, for example, that track cell phone usage, anonymized, but tracked cell phone usage in incredible detail, and track people moving in vehicles. Would not be hard to match that data to the automatic number plate recognition data and have a record of where everybody is at any given time that might be of value to private investigators investigating things or whatever it might be. And of course, the problem here, you and I have talked at length about this, and I ramble to anyone that will listen to me about this.

Brett Roberts:
So much of this protecting people relies on the goodwill of technology folks who I think have demonstrated amply over the last couple of decades that maybe they shouldn’t always be expected to be demonstrating goodwill. And these things move faster than the legislation to control them does. So I think it’s really important, more important than ever and getting more important by the day that we have folks in government that really are thinking about this stuff, that are putting legislation in place, good legislation, to protect this data and get rid of this data as and when it’s not needed anymore. It’s a critical, critical thing.

Paul Spain:
One of the things that we’ve talked about is what foodstuffs have been doing with their facial recognition in supermarkets. And look, one of the things around the way that they operate, they’ve got images of, you know, people that have, that have caused problems, shall we say, in their supermarkets. And so they will, you know, take video footage or grab a still frame, you know, compare it against that, that. But if you’re going walking into the supermarket and you’re not recognized as being on the list, then they’re dumping the data really quickly. That was one of the things that I thought was on the good side, the way that they’re making sure they’re not sitting on that data inappropriately. Now, there are other sort of fish hooks and complexities there, however, such as when the AI looks at somebody who is not an offender and actually thinks that they are and so on. But there are ways to steer and lean these things in a better direction than what they could be. And having the dialogue on that and encouraging governments to make good decisions, I think is pretty important.

Paul Spain:
And hopefully our, know, privacy commissioner, you know, gets more and more involved in, in these things and we don’t just sort of see an ongoing erosion of, you know, privacy.

Brett Roberts:
I think you, you raised some really good points there. I think the machine vision is, is technology. That’s always amazed me and terrified me simultaneously, I think Aura have done a fairly good job of, of walking a very difficult middle line around Some of the stuff, I think, you know, machine vision is only as good as the data that’s trained on. We’ve seen that there can be real issues with false positives. There’s been a few pretty bad stories in the press around some of that stuff. You know, Absolutely. There will be more and more of it used. You know, we just need to expect that.

Brett Roberts:
So I think it’s important that we hold the folks that are using these technologies, that we hold their feet to the fire to make sure things go well and when they don’t go well, that, you know, those things are. The issues are remedied quickly. I think Privacy Commissioner plays a huge role in this. And, you know, there’s some pretty meaty stories from the past. You know, for example, New Zealand police four or five years ago, some folks playing around with Clearview AI hadn’t told their superiors that they were doing that. And as far as I’m aware, they certainly hadn’t told the Privacy Commissioner because they were doing that. So, yeah, I think we need very, very clear guidelines around these things and methods to rein in technologies that start to do things to or for people that may have unintended consequences. And again, one of the challenges here is legislation moves at a glacial speed while the technology races hit at a digital speed.

Brett Roberts:
Juxtaposition’s always been challenged, right? Always will be.

Paul Spain:
Yeah. Look, these things are really hard to get right. And I was, I was hearing something just, just just briefly this morning around a government entity. And I haven’t got or looked into the full details, but they, without a, without a, a full warrant, had seized at least one, one mobile device. And I think they got the, they got the pushback on, on having done that and possibly ruined the court case that they were involved in. So we do have some reasonably good precedent around, you know, certain areas where, you know, we need to draw some lines. But of course, those things. And those lines will, will move and change in time as, as I’m sure.

Brett Roberts:
The folks from the, from the GCSB and SIS could tell all sorts of interesting stories around where those things collide. Right. Where, yeah, national security intersects with someone’s right to privacy. Yeah. Digital devices, etc. Yeah.

Paul Spain:
Now on this, this, this sort of front, it’s, it’s been announced that the, the government are testing out technology that will effectively help them catch those who are evading court fines by, you know, clamping and seizing their vehicle. Vehicles. This, this was, you know, has, has, has been announced and Paul Goldsmith, the Justice Minister, said bailiffs are trialing handheld devices that can scan number plates of, of parked cars and then it can determine whether the owners have overdue court fines or reparations and and so on. Now this does sound a little bit like a kind of a thin end of the wedge. Cause you try something like that and it’s like, oh, that doesn’t actually work very well with scanning. So actually the next recommendation is, well, you should be scanning everybody all the time and then flagging the car and having a dot on a map somewhere and you can go and get these people. And I sort of see both sides of it because there’s a real efficiency to that. But you know, then, yeah, what situation do you end up with? And as you mentioned, Brett, there are ways of doing that, you know, that are probably quite legal for private individuals or entities to invest in.

Paul Spain:
But we need the government to lead the way in fair and appropriate manners. And I’m not saying this one isn’t, but, but I think it’s just a cause for us to be looking in a little bit closer and thinking about the approach.

Brett Roberts:
Yeah, look, I think the challenge at one ended scale is I can absolutely see if someone owes many, many thousands of dollars in fines and has refused to pay them for a very long period of time, then clamping their car or seizing their car kind of makes sense. Right. For someone who, you know, doesn’t owe that much or maybe can’t afford to pay and their vehicle is their lifeline to employment. What’s the story there? I’ve always liked the thing that they do is Sweden or Finland where when they levy speeding fines, they do it on the basis of your salary, they charge you a percentage of your salary. So maybe we need to do something similar there. I think there’s a real danger of overreach here. And I think the other thing too is, and I know there are millions and millions, probably hundreds of millions of dollars in outstanding fines that need to do something about it. But I sometimes look at these things and think that, you know, from a return on effort perspective, are we putting as much effort into fighting white collar crime? We’re probably not, you know, are we investing as much effort into, you know, unpaid taxation? Should we be clamping the wheels of Porsche drivers whose companies have not paid their tax? You know, where do we, where do we draw the line? Right, yeah, I don’t know.

Brett Roberts:
I don’t think there’s any simple answer to it. But I do think it needs to be. And again, you know, the legislation needs to be in place to make sure these things do not run rampant.

Paul Spain:
Yep. And look, another, another area that’s been in the news in the last few days was around illegal dirt biking happening in residential areas around Christchurch.

Brett Roberts:
Yes.

Paul Spain:
Trying to work out, you know, how do you, how do you, you know, how do you police that? And, you know, of course anyone in Auckland is, you know, pretty familiar with the police, the helicopters that sort of seem to fly over probably some areas more than, more than others. And, you know, there have been people that have, you know, complained being kept up at night by them and whatnot. But yeah, the suggestion that was coming out of the Christchurch resident was, well, you know, those might not be up to needing a police helicopter, but maybe, maybe a drone going up to, you know, track them down and follow them. So, yeah, these sort of stories of the use of tech to, you know, hopefully, you know, produce, you know, a better society are going to keep coming and it’s how we sort of steer through that. So we, we lean much more on the, on the good side than, than, than the bad side because, yeah, I mean, technology itself, I guess is, is, is, isn’t good or bad. But yeah, if we get the, the uses a little bit off, then then we might suffer.

Brett Roberts:
There’s a great documentary on Netflix. It’s a documentary series and one of them is about the embedding of artificial intelligence into weaponry. And it is a very sobering and terrifying watch. It really is. And again, you know, that just these things are always a continuum. You know, using drones to patrol people and potentially identify people that are using dirt bikes in a way that they shouldn’t be in are illegal. Absolutely fine thing to do. Flying drones over citizens day in and day out, tracking what people are doing and matching it with car data on that sort of thing.

Brett Roberts:
Probably not a good thing. And I think the danger always with these things is people with good intentions accidentally doing things that aren’t necessarily good with them. And that’s why I think we need good guide rails, good legislation. We need folks with power to do things about things. If and when things go wrong, then that’s a critical part of it. Yep.

Paul Spain:
Now onto, we’ll change tack a little bit. Starlink. Look, you know, I think I’ve spoken to lots and lots of people who just love having Starlink as a connectivity option. It’s. Yeah, it really has changed the landscape and you know, especially some parts of the world where connectivity has been very limited. But, you know, also in rural areas, it creates a bit of competition for, you know, for our local Kind of telcos and you know, those providing the rail wisps, the wireless Internet service providers offering WI fi and the. Yeah, the sort of benefits, I guess, out at sea and so on from a connectivity perspective, lots of good stuff there. But last week we saw the flip side to it Friday when Starlink connectivity seemed to fail for, you know, a bunch of people for, you know, a couple of hours or more.

Paul Spain:
And there was still reports coming in sort of 24 hours later of, of people having issues. I’m not sure whether that was necessarily true or not or if it was more edge cases and people needed to reset their devices and so on. But that sort of impact that you get when you’re very reliant on one company is potentially quite disastrous, I think. What are your thoughts on this one?

Brett Roberts:
There’s a few things in here, right. So one is it’s an incredible technology and as anyone that knows, I’m going to tell I’m not a big Elon Musk fan, but you cannot help but admire incredible job they’ve done of, you know, launching all these satellites into space to, to build the Starlink network. Right. Astronomers hate them, obviously, and they terrible light pollution, et cetera and night pollution. Technology itself is quite incredible. The fact that it’s effectively under the control one, as noted previously, ketamine addicted billionaire, I don’t think is a good thing at all. So I think the fact that it’s under the control of one, one particular human being, not the most stable of human beings, I don’t think that’s a good thing. There are competitors.

Brett Roberts:
I know Jeff Bezos is doing the same sort of thing and I believe there’s a UK company and I believe the Chinese are also looking to do it. So there will be competition, which is good and important. But I think, I think the fact that, and it doesn’t matter who really controls the company, but I think the fact that such an important technology being used, certainly in, you know, important use cases, like, for example, you know, connecting Ukrainian soldiers with intelligence and, you know, important military information, being able to be turned on and off by somebody or even just due to technical fault, is a single point failure issue that I think really does need to be, really does need to be addressed. I live in an area where lots of folks use Starlink. They know a lot of folks around Estonia 5. They absolutely rave about it. It’s clearly a good service. But again, all of that, all of that power, all of that control vested into the hands of a single company or a single individual, I think is a accident or tragedy or something waiting to happen.

Brett Roberts:
Yeah, so more competition in that space, the quicker is, you know, quickly it’s a good thing.

Paul Spain:
Yeah, look, I think it’ll be interesting to see how it plays out. I was reading some details on how many satellites that they, they’ve, that they’ve you know, put up and the numbers are you know, pretty, pretty, pretty astounding and it’s going to take I think quite some time for anybody else to you know, to be able to agree, match them in terms of you know, coverage and performance and, and, and so on. So yeah, 9200 Starlink satellites have been, have been put, incredible put into orbit. And when I think of. So that’s since 2019 and if we look back to 2019 I think the number of satellites in total that were in orbit at that time is probably sub 5,000. So because of, you know, because of Starlink we, we’ve seen this, you know, what’s you know, somewhat exponential kind of growth in the number of satellites. Now I think 7900 of those Starlink satellites are still operational. So they do, you know, deorbit, I’m sure a mix of ones that have, you know, been formally decommissioned and those which it hasn’t gone so well and they’ve, you know, they’ve come back into the atmosphere sort of on a more unplanned type of basis.

Paul Spain:
So yeah, it is pretty phenomenal and yeah over 6 million subscribers. So it’s pretty formidable what they’ve managed to launch. So yeah, we will see how that goes with competition. I imagine there’s room here for a number of players but it’s a really, really big investment and I think some people are wondering, well what’s the role of Rocket Lab in the future of launching this type of technology? They’ve been touting what they call the flat alight. So an ability to launch a whole lot of flat satellites and so on as a service. So yeah, these things will get easier but we’ve probably still got a little way ahead with there being one predominant supplier in space based comms for the average consumer.

Brett Roberts:
And the other thing of course just adding to all of this is irrespective of who might be providing the bandwidth, the impact this will have on the classic terrestrial telco companies. You know, none of those folks must be happy every time they see another Starling satellite guard. Right?

Paul Spain:
Well yeah, it’s interesting because you know, talking to the telcos they very much I guess have had some warning the technology was coming and in fact you know, they are leveraging this type of, you know, technology when there’s disasters and other things for emergencies, you know, bring cell sites online and the like and you know, of course providing that sort of fill in terms of coverage and I think that aspect of being able to sort of fill in, you know, coverage wise is just something that will just become the norm for all of us and you know, we’ll get used to that just being the way it is. So you know, I guess it’s not just competitions so.

Brett Roberts:
No, no, true, that’s true.

Paul Spain:
Yep, yep. Now, yeah, yeah, we’ve, we’ve had obviously one NZ sort of progressing on that front with this sort of satellite side of things and AST Space Mobile, um, you know they’re the, they’re the compete and um, actually they’ve been, they, they’ve been in news recently because they’ve been able to place a call. So their tech is out there competing but theirs is focused on the mobile to satellite side. So yeah, I think what do we see? It was, yeah just in the last week that in conjunction with AT&T in the US they completed their first VoLTE call. So voiceover LTE or 4G connection. So you know, more than just texting, which is mostly what we’ve seen so far but already testing that out on a calling basis. Now for that to really work AST Space Mobile is going to need to get a lot of satellites out there but they’re focusing on doing it with a lot less satellites that are much, much bigger. So that’s going to be quite fascinating to you know, to see how quickly they can actually get to market and you know, how quickly I guess our other telcos in New Zealand will get access to, you know, texting and calling type service.

Paul Spain:
I guess at the moment it’s a little bit of a wait and see. Okay, now onto, onto other topics. It’s interesting, I know you’re not a big Trump fan but he’s, he’s drawing more headlines. He has unveiled an AI plan that aims to clamp down on regulations and bias. Was the, the headline that, that the story was published under by, by the BBC and it’s a 28 page AI action plan and outlines, you know, a whole, you know, suite of, of policy actions. I think about 90 there and, and you know, big part here for the US is that they seek to do very, very well because so much of the AI is coming out of the US and you know, I guess what, you know what, what we’re seeing there is that you Know, Trump and US Government are looking to, well, how do they, you know, keep the, keep the foot down on this and, you know, see the U.S. you know, accelerate what the revenue opportunities are for the country. And you know, there’s a whole bunch of aspects to that, but I think it’s down to the MAGA thing, the make America Great thing.

Paul Spain:
They want to, you know, take advantage as much as they can and in theory benefit their country. Now, I’m not sure exactly how this, how this sort of stuff plays out and who’s behind it.

Brett Roberts:
Right.

Paul Spain:
So who are the people that are putting this together? Obviously, you know, Trump’s the guy sort of waving the flag as the president on this stuff, but there’s going to be a big bunch of people behind that. And whether that’s the billionaires or, you know, really smart people that have all of our best interest in mind, you know, I’m always a little bit skeptical of these sort of things. Right?

Brett Roberts:
Yeah, look, there’s a whole bunch of things going on here. I mean, this is an arms race, right? It is an AI arms race. And if you kind of step right back from it, the two things that I see happening right now is that there’s a huge amount of innovation and invention happening in the Western world, predominantly in the United States, around artificial intelligence, some quite incredible stuff. Levels of investment that we’ve literally never seen before. Hundreds and hundreds and hundreds of billions of US Dollars going into this. Those driving up stock prices of those companies. Nvidia, Microsoft, Google, Meta, etc, propping up the standard and PA 500 in the NASDAQ. You know, if Nvidia sneezes, I think the stock market’s going to catch bubonic plague.

Brett Roberts:
So there’s, you know, from the US perspective, this is a, an economic opportunity that they want to, that they want to win. And then you see what China are doing, which I think is a very clever strategy of effectively attempting to demonetize the value of, of these AI things. So, you know, you’ll get Sam Altman come out with some amazing announcement about something absolutely incredible that’s got some quadrillion dollars to put together and a bunch of Chinese researchers have put something together that’s almost as good and good enough and they’ll release it, you know, as an open source model just to pick away at it. So I think that’s going to be an interesting race because at some point someone’s going to have to pay the piper with regard to all of the investment that’s gone into this. Those investors will Want to see a return on that investment. Every one of these companies at the moment is burning billions and not making anything in return. Investors will speculate for a while, but not forever. So I think there’s quite an interesting cat and mouse game going on between the east and the west at the moment that I’m finding quite intriguing.

Brett Roberts:
My personal view is that I think it is a bubble that’s overhyped and it will pop at some stage. I just don’t think that the economics stack up because they currently stand. And so who’s in behind this? I think it’s a pretty safe bet that it’s the likes of Peter Thiel, Sam Altman, Zuckerberg, and all of those folks will be the ones that are prompting this and say Trump’s out in front waving the flag. This really is an arms race, and I think it should be viewed as such. And as I said, I think it’s. What I find interesting is the Chinese aren’t. Are playing the same game, they’re playing a very different game. And it will be interesting to see what happens if, you know, these companies are spending hundreds of billions of dollars to, to build these incredible models, and the Chinese keep coming out with things that are not too bad with an open source, in an open source kind of way, be very hard to monetize.

Brett Roberts:
There’s a great analogy that I’ve heard Scott Galloway use on this, which is about where the value and the AI boom will land. And he compares it to the aviation industry. If you think about the incredible innovation and investment that went into creating modern jet engines, modern avionics systems, these incredible aircraft that, you know, enable you to buy a ticket to get to the other side of the world for, you know, for $1,000 if you shop on the right day. You know, the value for that innovation in that industry landed in the hands of the travelers, you know, Boeing, who, who would invest in an aircraft company, who would invest in an airline. Right? How do you make a small fortune in the airline industry and start with a large one? So all of that innovation, all that incredible invention, a lot of, again, which came out of the west, really has not benefited the owners and the shareholders of those companies. It’s landed with the users of their technology. And I actually think AMI will prove to be the same. A lot of companies are going to have a very, very bad time seeing their billions evaporate as they end, or the Chinese or others put technology into the hands of folks that people will like and use but might not pay for or not Pay very much.

Paul Spain:
Yeah, Look, I think we want to kind of be stepping back, looking at it all through a New Zealand context too, don’t we? Because you know, what you state is there’s very genuine risks. There’s obviously lots of companies that are trying to ride the wave of AI and some will get that right and do very, very well, other ones will fail. But yeah, it is those big players ultimately whose technology probably most of us are going to be going to be using. And you know, I’m very, very hopeful that, you know, as a country we can find those right areas to plug in to leverage this technology that, yeah, ultimately a lot of those companies may well not do very well. I think some of the companies will try and ride it from, you know, multiple perspectives. And that’s maybe where Microsoft are in terms of, you know, their. They’re really looking to be across the value chain and you can see, I mean, if their current share price is anything to go by, and I personally haven’t held a Microsoft share for quite a long time, but it suggests people believe that maybe they’re getting it right on this front. But of course there are so many opportunities at home.

Paul Spain:
Companies like Xero continue to sort of leverage differing aspects of machine learning and AI in ways that improve things for their customers and then they benefit because people are willing to pay their subscriptions. We’ve got, I think, a whole bunch of differing use cases in New Zealand, be it individual organizations who are finding the right ways to leverage AI within their business, within their workflow. You’ve got JD Trask, who’s been on the podcast before, who runs Raygun Wellington. They’ve launched Auto Hive, right, which looks like a really interesting set. I’ve had a set of tools I’ve had a little bit of a look at, like, look at that. It’s probably still reasonably early days, but it’s exciting to see that we’re not all asleep at the wheel on these things. But it’s really, really hard to get it right. So it’s kind of hard to know who’s really going to win.

Paul Spain:
But yeah, I think what’s going to.

Brett Roberts:
Be interesting here is I think anybody that’s building an AI business that sits directly at top some of the large models, you know, whether it’s OpenAI or whoever it might be, I think they’re going to have a bad time because that, you know, that boundary will just move out. Those, Those companies, open AIs of this world will just start to provide those services that these other little companies are building. So I think it will be very difficult to maintain a competitive moat if you are relying directly on, on the likes of OpenAI. And of course the other challenge you have there is that OpenAI could turn off access or quadruple the price of access. And so that from an economic model perspective it’s very difficult to ensure surety of supply or surety of access to those things. I think companies that are building their own models using their own data, I think they definitely, that’s a different case in Thailand. But yeah, I, I think it, it will be interesting to see just how this all plays out. There’s going to be a lot of bodies lying around after the, after things go bang.

Brett Roberts:
And on the subject of Microsoft, I think they, they’re playing quite an interesting kind of bit hedging strategy. But it also, you know, and they are, you know, it’s a very, very savvy company that I would never beat against. But it’s interesting to see that they are, you know, automatically moving people onto new Microsoft 365 plans with Copilot without asking him.

Paul Spain:
Right, that’s a bit sneaky or a lot sneaky.

Brett Roberts:
Yeah, that is sneaky. And, and to get yourself downgraded back to your previous plan, you have to pick up the phone and call up and that to me smacks of desperation. So I think, I think there are a lot of folks that have invested very, very heavily in this that do want to see or will want to see a return on investment very, very soon. And I’m not 100% sure that we’re going to be seeing that to the scale that they, that they require. So I think we’ve got some very interesting months ahead of us. I think the end of this year, you know, and there’s the whole thing with OpenAI changing their business model at the end of this year. There’s, there’s a lot of speed bumps to drive over between now and then. It’s going to be something that changes by the hour.

Paul Spain:
Yeah. One of the business use cases for AI is using sort of varying forms of sort of surveillance data and data that’s captured about individuals to create somewhat dynamic based pricing in areas like insurance and so on that AI’s got a whole lot of info and then, you know, potentially adjust pricing. But I see that two Democrats in the, in the US have, have, you know, putting to Congress a law that would basically block firms from jacking up prices, algorithmic pricing, you know, based on, based on all the data that they’ve got. And yeah, Algorithms being able to go in there and make those adjustments now. Yeah, I’m not sure how this will play out, but I think it’s probably, if nothing else, just a reminder to us that, you know, a use case that looks good for a business or looks good for, you know, government entity isn’t necessarily going to be something that people are going to smile about and, you know, we’re going to have to, you know, maybe challenge some of these situations. So, yeah, I’m not sure if that one will probably, you know, cross the line.

Brett Roberts:
I don’t rate the chances. Right. Yeah, airlines do it. You know, there’s the apocryphal story, which I’m sure is true of, you know, your Uber costs more if your battery’s almost flat. You know, algorithmic pricing has been around for a long time and, and I do not see, I do not see any hope of it being reined in or, or, or eradicated. You know, when you look at, you know, the. How the light. And Uber’s a great example, you know, just, just how they use what they know about you and your circumstances and your location, etcetera, to, to frame up pricing on the fly, that’s effectively their business model.

Brett Roberts:
And, and, you know, I just don’t see lawmakers being able to drag that over the line.

Paul Spain:
Yeah, I mean, yeah, I’m kind of curious, you know, what’s the right sort of balance? Right. Because, yeah, some of those things we’re totally used to, they’re totally normalized, but there are probably some extremes we’d be pretty uncomfortable with if you were to drill into all the data that they’ve got and some of the things that they could, could choose to do. Now, last up before we finish, just thought I’d mention upcoming event next week, Revved, is taking place in Auckland and it’s sort of, I guess, a mix of sort of leaders from sort of business and tech worlds, but, you know, quite a few names that folks would recognise and probably enjoy hearing from across the, you know, the tech sector. A bunch of folks in the AI world. So we’ve got Carolyn Rainsford, who heads up Google in, in New Zealand. We’ve got Dr. Masa McCauley, who is, is in the AI world, amongst, amongst other things, Mark Lawrence, generative AI consultant. We’ve got Andy, Andy Higgs, who many may know, who’s now with Digital Identity nz.

Paul Spain:
Vanessa Sorensen from Microsoft, Sir Ian Taylor, who we’ve had on the New Zealand business podcast and, you know, a whole bunch more. So for those that are you know, feeling like getting out of the office and you know, a chance to networking network and learn some things from some pretty smart people. People both those on on stage and and probably other attendees might well be worth with your time popping along. I’m looking forward to get getting down there Thursday next week. So feel free to you know, come and say hi if, if you see see me about there.

Brett Roberts:
Looks like a good event. I have to say that’s an awesome lineup of. Of speakers. Yeah.

Paul Spain:
Yeah. Oh, yep. Some some really, really great people in there and that have only just mentioned a, a few that were put in front of me, but there’s a bunch more. So yeah, worth googling it and going and having a look at the details there. For revved on 7th of August, I think that’s us for this week. Anything else you wanted to add in, Brett?

Brett Roberts:
Oh, I think that was an extensive coverage of a lot of things going on in the digital world this week. That was good.

Paul Spain:
Yeah. Really really appreciate having your insights as always, Brett. Really, really great to catch up, hear some of your your thoughts on all these things going on in the world of tech.

Brett Roberts:
You know me, I have an opinion or two.

Paul Spain:
Fantastic. Oh well, thanks everyone for, for joining us. Big thank you of course to our show partners, Workday, Gorilla Technology, Spark, HP, 2degrees and One NZ. We’ll look forward to catching you next week on the next episode. Also make sure you’re checking out New Zealand business podcast. We’ve got got some new episodes coming up on that front and as is often the case, a lot of the trailblazers in the business world are also in the tech world. So yeah, I think you’ll find some interesting listens across there too. Alright, thanks everyone.

Paul Spain:
And thank you to Joe Allen, our producer here in the Podcast NZ studio for keeping everything ticking along and handling our remote recording today. Much appreciated as always and thanks again, Brett.

Brett Roberts:
Thanks, Paul. See you next time.

Paul Spain:
All right, cheers.