1 00:00:12,920 --> 00:00:15,480 Speaker 1: I'm Peter Griffin and welcome to the Business of Tech 2 00:00:15,520 --> 00:00:18,720 Speaker 1: powered by Two Degrees, a special episode I've put together 3 00:00:19,200 --> 00:00:22,400 Speaker 1: to try and unpack a really big tech related story 4 00:00:22,400 --> 00:00:26,040 Speaker 1: that played out this week, the launch of Amazon Web 5 00:00:26,079 --> 00:00:30,440 Speaker 1: Services New Zealand Data Center Region four years in the making. 6 00:00:30,600 --> 00:00:33,920 Speaker 1: It should have been a big milestone to celebrate, but 7 00:00:34,040 --> 00:00:36,960 Speaker 1: it turned into a train wreck for Amazon, which really 8 00:00:37,040 --> 00:00:40,400 Speaker 1: shot itself in the foot by refusing to explain why 9 00:00:40,440 --> 00:00:44,520 Speaker 1: it had seemingly pivoted from building its own shiny, expensive 10 00:00:44,640 --> 00:00:49,400 Speaker 1: data centers to leasing space in existing ones and what 11 00:00:49,479 --> 00:00:52,159 Speaker 1: that meant for the seven point five billion dollar investment 12 00:00:52,200 --> 00:00:55,160 Speaker 1: over fifteen years it promised back in twenty twenty one, 13 00:00:55,600 --> 00:01:00,800 Speaker 1: and the thousand jobs it expected to create AWS, well, 14 00:01:00,840 --> 00:01:02,720 Speaker 1: that was just an option on the table. Never mind, 15 00:01:03,040 --> 00:01:07,440 Speaker 1: we've done something different. Journalists called bs on that the 16 00:01:07,560 --> 00:01:12,199 Speaker 1: media coverage was wall to wall brutal. It's perhaps best 17 00:01:12,280 --> 00:01:16,120 Speaker 1: summed up by Heather Dupless Allen's news Talk ZB interview 18 00:01:16,560 --> 00:01:20,920 Speaker 1: with AW West New Zealand country manager Manuel Bonnett on Tuesday. 19 00:01:21,600 --> 00:01:22,480 Speaker 1: Here's a clip from it. 20 00:01:23,120 --> 00:01:25,520 Speaker 2: Have you built any data centers? 21 00:01:25,920 --> 00:01:28,640 Speaker 3: So a region is a cluster of data centers, and 22 00:01:28,680 --> 00:01:32,720 Speaker 3: of course we have chosen the locations that we choose 23 00:01:32,800 --> 00:01:37,200 Speaker 3: to build out our availability zones. The region consists of 24 00:01:37,280 --> 00:01:40,720 Speaker 3: multiple availability zones. We have three of these availability zones 25 00:01:40,720 --> 00:01:42,760 Speaker 3: here and of course that involves data centers. 26 00:01:42,880 --> 00:01:44,880 Speaker 2: Cool and how many of these have you actually built 27 00:01:44,920 --> 00:01:46,520 Speaker 2: as opposed to renting or leasing. 28 00:01:47,440 --> 00:01:50,400 Speaker 3: So we are choosing facilities based on the needs that 29 00:01:50,480 --> 00:01:54,280 Speaker 3: we have, and we have thirty eight of these regions globally. 30 00:01:54,280 --> 00:01:56,639 Speaker 2: Exactly what you're being worried about this though, aren't you, Manuel, 31 00:01:56,640 --> 00:01:57,800 Speaker 2: because you haven't built any have you? 32 00:01:58,600 --> 00:02:01,160 Speaker 3: I'm telling you we chose the locations that are the 33 00:02:01,160 --> 00:02:01,560 Speaker 3: best looking. 34 00:02:01,680 --> 00:02:03,520 Speaker 2: What it happened? Why didn't you guys build it? It 35 00:02:03,520 --> 00:02:05,120 Speaker 2: wasn't that part of the announcement with just in the 36 00:02:05,120 --> 00:02:06,960 Speaker 2: back in twenty twenty one that you were going to build. 37 00:02:06,720 --> 00:02:10,480 Speaker 3: Some Well, we always take a long term view and 38 00:02:10,520 --> 00:02:12,399 Speaker 3: as part of that we look at all the options 39 00:02:12,480 --> 00:02:15,239 Speaker 3: and I'm super happy that we found the right locations 40 00:02:15,240 --> 00:02:17,040 Speaker 3: for the region to be live as of today. 41 00:02:17,160 --> 00:02:19,079 Speaker 2: Okay, what happened to the west Aokland one where you 42 00:02:19,200 --> 00:02:21,160 Speaker 2: flattened it? You got rid of the wetland? Are you 43 00:02:21,200 --> 00:02:22,280 Speaker 2: going to build anything there? 44 00:02:23,280 --> 00:02:23,560 Speaker 4: Well? 45 00:02:23,639 --> 00:02:25,680 Speaker 3: I mean, for now we are announcing that the region 46 00:02:25,760 --> 00:02:29,160 Speaker 3: is live, so the availability zones at the manual. 47 00:02:29,240 --> 00:02:32,280 Speaker 2: I just need to tell you, mate, listen. This audience 48 00:02:32,320 --> 00:02:34,320 Speaker 2: that listens to this is really smart and they can 49 00:02:34,360 --> 00:02:36,359 Speaker 2: hear you not answering the questions. So you can keep 50 00:02:36,400 --> 00:02:38,040 Speaker 2: on with that strategy if you want to. But I 51 00:02:38,160 --> 00:02:40,840 Speaker 2: need to warn you. We know you're not answering my question. 52 00:02:40,880 --> 00:02:42,880 Speaker 2: I'm going to try it again. What are you doing 53 00:02:42,880 --> 00:02:44,520 Speaker 2: with the West Aalkland site? Are you going to build 54 00:02:44,520 --> 00:02:45,120 Speaker 2: anything there? 55 00:02:45,919 --> 00:02:49,240 Speaker 3: I can't comment on the site. I'm basically I'm announcing 56 00:02:49,280 --> 00:02:51,120 Speaker 3: the launch of the region and the site, as you 57 00:02:51,160 --> 00:02:52,280 Speaker 3: can tell, is not part of it. 58 00:02:52,400 --> 00:02:55,360 Speaker 2: Okay, thanks mate, appreciate it. Go well. Manuel Bonnette, New 59 00:02:55,440 --> 00:02:58,640 Speaker 2: Zealand Country Manager, Amazon Web Services. Oh my gosh, guys, 60 00:02:59,480 --> 00:03:01,919 Speaker 2: how bummed to you about this announcement. I thought when 61 00:03:01,919 --> 00:03:04,400 Speaker 2: I heard luxon on with Hosking this morning, I thought 62 00:03:04,440 --> 00:03:05,120 Speaker 2: this was awesome. 63 00:03:05,720 --> 00:03:09,040 Speaker 1: Amazon really didn't understand what they were getting themselves into 64 00:03:09,160 --> 00:03:12,840 Speaker 1: when they dreamt up the narrative behind this launch. The 65 00:03:12,919 --> 00:03:15,720 Speaker 1: fact that the Prime Minister then jumped on it for 66 00:03:15,760 --> 00:03:19,800 Speaker 1: a political capital just made it so much worse, which 67 00:03:19,840 --> 00:03:23,239 Speaker 1: is a shame because going live with a new hyperscale 68 00:03:23,480 --> 00:03:27,840 Speaker 1: data center region is a big development. AWS has made 69 00:03:27,960 --> 00:03:33,320 Speaker 1: a significant investment in cloud computing and is committing to 70 00:03:33,440 --> 00:03:36,720 Speaker 1: having a significant presence in New Zealand. That's a good thing. 71 00:03:37,280 --> 00:03:40,080 Speaker 1: So I thought i'd get a true cloud computing industry 72 00:03:40,160 --> 00:03:44,480 Speaker 1: expert on to unpack what went on here, or what's 73 00:03:44,560 --> 00:03:47,640 Speaker 1: probably gone on. It's hard to know exactly because everything 74 00:03:47,680 --> 00:03:51,520 Speaker 1: is shrouded in secrecy when it comes to AWS in 75 00:03:51,560 --> 00:03:55,240 Speaker 1: the name of security apparently. Ben Keeps is a christ 76 00:03:55,360 --> 00:03:59,600 Speaker 1: Church businessman, board director and cloud industry analyst who was 77 00:03:59,640 --> 00:04:03,040 Speaker 1: one of the earliest New Zealand proponents of this migration 78 00:04:03,160 --> 00:04:06,200 Speaker 1: to the cloud many businesses have been on. He knows 79 00:04:06,240 --> 00:04:09,480 Speaker 1: the AWS tech stack in depth, and he knows a 80 00:04:09,520 --> 00:04:13,640 Speaker 1: lot about the economics of hyperscale data centers. So he's 81 00:04:13,640 --> 00:04:17,280 Speaker 1: been with his candid take on what went down this week. 82 00:04:17,640 --> 00:04:20,120 Speaker 1: Ben Keeps, Welcome to the business of tech. How are 83 00:04:20,160 --> 00:04:20,520 Speaker 1: you doing. 84 00:04:20,640 --> 00:04:22,280 Speaker 4: I'm doing well. Thanks for inviting me. 85 00:04:22,560 --> 00:04:26,800 Speaker 1: Well, what a debarcle I mean this week with AWS 86 00:04:26,800 --> 00:04:31,239 Speaker 1: probably one of the worst PI disasters in the tech industry, definitely, 87 00:04:31,240 --> 00:04:34,520 Speaker 1: But in terms of any big corporation in recent memory 88 00:04:34,720 --> 00:04:38,679 Speaker 1: in New Zealand, what's your take on this? What really 89 00:04:38,680 --> 00:04:39,680 Speaker 1: went wrong? This week. 90 00:04:39,800 --> 00:04:42,839 Speaker 4: Yeah, I mean, you're right, it was absolutely abysmal. There's 91 00:04:42,880 --> 00:04:46,560 Speaker 4: two stories here, right. The first story is one about investment, 92 00:04:47,520 --> 00:04:54,120 Speaker 4: enzting infrastructure politics frankly, and that was a total debarcle 93 00:04:54,240 --> 00:04:59,680 Speaker 4: and it was because of political expediency on the politician 94 00:04:59,760 --> 00:05:05,599 Speaker 4: side and frankly immaturity, surprising immaturity on the vendor side. 95 00:05:06,320 --> 00:05:08,600 Speaker 4: The second part of the story, which is actually impactful 96 00:05:08,720 --> 00:05:13,640 Speaker 4: and got completely lost, is that ADS is actually setting 97 00:05:13,720 --> 00:05:16,359 Speaker 4: up or has set up a region here in New Zealand. 98 00:05:17,000 --> 00:05:19,599 Speaker 4: And that answers a bunch of questions that people have 99 00:05:19,720 --> 00:05:22,559 Speaker 4: had for going on twenty years now. 100 00:05:22,720 --> 00:05:26,200 Speaker 1: So let's unpack these things. Let's go all the way 101 00:05:26,240 --> 00:05:30,440 Speaker 1: back actually to twenty twenty one, the day after AWS 102 00:05:30,480 --> 00:05:33,320 Speaker 1: made that announcement, the initial announcement back in the depths 103 00:05:33,320 --> 00:05:36,119 Speaker 1: of COVID that they were setting up a data center 104 00:05:36,200 --> 00:05:39,560 Speaker 1: region in New Zealand. This is what you wrote in 105 00:05:39,600 --> 00:05:42,000 Speaker 1: the New Zealand Herald. The devil is in the detail 106 00:05:42,040 --> 00:05:44,520 Speaker 1: and apart from the big number seven point five billion 107 00:05:44,680 --> 00:05:48,400 Speaker 1: investment and one thousand jobs, there's a little detail in 108 00:05:48,440 --> 00:05:52,160 Speaker 1: the announcement, but details aside. This is exciting and further 109 00:05:52,200 --> 00:05:55,880 Speaker 1: puts New Zealand on the map as a credible technology nation. 110 00:05:56,040 --> 00:05:58,239 Speaker 1: That was a full four years ago. And what's striking 111 00:05:58,240 --> 00:06:01,919 Speaker 1: about that is those same figures are exactly what was announced, 112 00:06:02,400 --> 00:06:05,279 Speaker 1: you know, on Monday at Amazon's Cloud Day. And I 113 00:06:05,279 --> 00:06:08,280 Speaker 1: think that was the root of the problem, was that 114 00:06:08,320 --> 00:06:12,080 Speaker 1: there was no advancement on this or no transparency into 115 00:06:12,800 --> 00:06:16,320 Speaker 1: the numbers. How the data centers, because there's three of 116 00:06:16,360 --> 00:06:20,040 Speaker 1: them in thew Auckland region are configured. Whether you know 117 00:06:20,120 --> 00:06:23,279 Speaker 1: Amazon is building its own, we know it's it's not 118 00:06:23,400 --> 00:06:26,040 Speaker 1: at the moment, it's it's co locating. Do you think 119 00:06:26,080 --> 00:06:28,440 Speaker 1: that's where the root of the problem started here before 120 00:06:28,440 --> 00:06:31,960 Speaker 1: the politics came into it. Just that lack of transparency, 121 00:06:32,720 --> 00:06:34,680 Speaker 1: which key with is you know, straight up people that's 122 00:06:34,720 --> 00:06:37,680 Speaker 1: sort of what they expect in dealing with companies. 123 00:06:38,320 --> 00:06:41,200 Speaker 4: Yeah, so seven point five billion back in twenty twenty 124 00:06:41,200 --> 00:06:45,560 Speaker 4: one was a terrible number because power, you know, buying 125 00:06:45,600 --> 00:06:49,360 Speaker 4: smoke over the construction crew, you know, the driver that 126 00:06:49,480 --> 00:06:53,440 Speaker 4: drives the you know, the aws CEO to cut a ribbon. 127 00:06:53,880 --> 00:06:56,200 Speaker 4: I mean, it's ridiculous. It was a ridiculous number then 128 00:06:56,440 --> 00:07:00,279 Speaker 4: and it's an even more ridiculous number now. However, the 129 00:07:00,320 --> 00:07:03,800 Speaker 4: thing that's really stark is that if AS had in 130 00:07:03,880 --> 00:07:08,040 Speaker 4: fact built three of its own data centers, the investment 131 00:07:08,160 --> 00:07:10,320 Speaker 4: would have been would have been massive. I don't know, 132 00:07:10,440 --> 00:07:12,040 Speaker 4: you know, it wouldn't have been seven point five billion, 133 00:07:12,040 --> 00:07:14,960 Speaker 4: but it would have been massive. The investment now is 134 00:07:15,880 --> 00:07:20,800 Speaker 4: significantly smaller because for all very good reasons, they are 135 00:07:20,880 --> 00:07:24,160 Speaker 4: embarking on a collocation strategy, which means that they're using 136 00:07:24,440 --> 00:07:29,320 Speaker 4: data centers that already exist. Obviously higher levels of capacity 137 00:07:29,360 --> 00:07:33,360 Speaker 4: within those data centers, but putting in a few extra servers, 138 00:07:33,440 --> 00:07:35,160 Speaker 4: or even lots and lots of extra servers, is a 139 00:07:35,200 --> 00:07:36,920 Speaker 4: hell of a lot cheaper than building an entire new 140 00:07:37,040 --> 00:07:41,720 Speaker 4: data center. So frankly, and the thing that really disappoints 141 00:07:41,760 --> 00:07:46,120 Speaker 4: me about the AOS launch, their part of it is 142 00:07:46,160 --> 00:07:49,040 Speaker 4: that to put out a statement that said seven point 143 00:07:49,080 --> 00:07:53,040 Speaker 4: five billion dollars a thousand jobs, when everyone knows the 144 00:07:53,120 --> 00:07:56,160 Speaker 4: data center is still just a hole in the ground 145 00:07:56,240 --> 00:08:00,160 Speaker 4: or not even and they are doing COLO is not 146 00:08:00,200 --> 00:08:03,760 Speaker 4: only is it duplicitus, but it actually doesn't recognize the 147 00:08:03,800 --> 00:08:07,480 Speaker 4: fact that we're not stupid, you know. And I've given 148 00:08:07,520 --> 00:08:10,600 Speaker 4: that feedback to my people with an AS. It was 149 00:08:10,680 --> 00:08:14,280 Speaker 4: a terrible launch. They didn't do themselves any fabors and 150 00:08:14,760 --> 00:08:18,920 Speaker 4: the really great thing about a region got lost because 151 00:08:19,440 --> 00:08:22,680 Speaker 4: they put out the release and obviously, you know, the 152 00:08:22,680 --> 00:08:25,320 Speaker 4: government decided this is a really good news story and 153 00:08:25,480 --> 00:08:28,000 Speaker 4: jumped onto it without being well briefed upon what it 154 00:08:28,040 --> 00:08:28,720 Speaker 4: actually meant. 155 00:08:29,000 --> 00:08:31,680 Speaker 1: Yeah, it is. There is a good story buried in there, 156 00:08:31,720 --> 00:08:34,120 Speaker 1: and we'll get to that. But just a bit more 157 00:08:34,160 --> 00:08:36,160 Speaker 1: under seven point five billion. I mean, you just have 158 00:08:36,240 --> 00:08:40,040 Speaker 1: to look at the AWS financial accounts over the last 159 00:08:40,200 --> 00:08:43,079 Speaker 1: few years. You know, typically a tech company, you know, 160 00:08:43,160 --> 00:08:46,520 Speaker 1: big tech, they usually spend about twelve and a half 161 00:08:46,520 --> 00:08:52,800 Speaker 1: percent off their sales, their revenue on capital expenditure. You know, 162 00:08:52,960 --> 00:08:56,000 Speaker 1: AWS did last year in New Zealand four hundred and 163 00:08:56,080 --> 00:08:59,720 Speaker 1: twenty five million dollars in revenue, really healthy business. But 164 00:08:59,760 --> 00:09:02,800 Speaker 1: when you look at their cap X, you know, purchases 165 00:09:02,840 --> 00:09:05,600 Speaker 1: of property and equipment, it's all in the you know, 166 00:09:05,960 --> 00:09:08,760 Speaker 1: sixteen million in twenty twenty one, eleven million in twenty 167 00:09:08,800 --> 00:09:11,760 Speaker 1: twenty two, eight point four million, and twenty twenty three, 168 00:09:12,400 --> 00:09:15,959 Speaker 1: it's clear that they're just not spending the big dollars 169 00:09:15,960 --> 00:09:18,400 Speaker 1: that they promised, which would have been to the tune 170 00:09:18,400 --> 00:09:20,679 Speaker 1: of five hundred million dollars a year. If you're looking 171 00:09:20,720 --> 00:09:24,760 Speaker 1: at seven point five billion over fifteen years. So they 172 00:09:25,559 --> 00:09:27,640 Speaker 1: they may have had these aspirations at one point, but 173 00:09:27,760 --> 00:09:31,319 Speaker 1: when those builds were off the table and they moved 174 00:09:31,559 --> 00:09:36,880 Speaker 1: to leasing arrangements, that probably cut the investment by half 175 00:09:36,920 --> 00:09:37,640 Speaker 1: at least. 176 00:09:37,600 --> 00:09:39,719 Speaker 4: Well well even beyond that. I mean, even if you 177 00:09:39,800 --> 00:09:42,360 Speaker 4: look at the original number with the building out data 178 00:09:42,360 --> 00:09:45,360 Speaker 4: center a strategy, So four hundred million revenue a year, 179 00:09:45,480 --> 00:09:49,200 Speaker 4: let's say you know they double that, triple that, quadruple that, 180 00:09:49,679 --> 00:09:51,680 Speaker 4: you know, a couple of billion dollars revenue a year. 181 00:09:52,760 --> 00:09:56,320 Speaker 4: You don't really spend seven point five billion dollars, you know, 182 00:09:56,440 --> 00:09:59,920 Speaker 4: building out on CAPEX, building that out in a very 183 00:10:00,080 --> 00:10:02,120 Speaker 4: finite market. I meant the end of the day, New 184 00:10:02,240 --> 00:10:05,800 Speaker 4: Zealand is always a small market for everything. And so 185 00:10:07,000 --> 00:10:08,440 Speaker 4: you know, that's why I said in twenty two one, 186 00:10:08,480 --> 00:10:10,960 Speaker 4: the devil is in the detail. I didn't believe that 187 00:10:11,040 --> 00:10:15,040 Speaker 4: the market opportunity here justified seven point five billion dollars. 188 00:10:15,679 --> 00:10:18,960 Speaker 4: The other, the other aspect that a few commentators have 189 00:10:19,080 --> 00:10:22,480 Speaker 4: sort of talked upon is that, you know, it was 190 00:10:22,720 --> 00:10:26,439 Speaker 4: it was touted as sort of corporate philanthropy and magnanimity 191 00:10:26,520 --> 00:10:30,120 Speaker 4: and how amazing as is investing in New Zealand and frankly, 192 00:10:31,080 --> 00:10:35,520 Speaker 4: you know, go capitalism there. They're a for profit organization. 193 00:10:36,040 --> 00:10:39,679 Speaker 4: They invest to make a return and so therefore there's 194 00:10:39,720 --> 00:10:43,640 Speaker 4: no philanthropy in there whatsoever. They look at a capex 195 00:10:43,800 --> 00:10:46,199 Speaker 4: and the payback on that. It's it's about net present 196 00:10:46,280 --> 00:10:49,480 Speaker 4: value and discounted cash flows and those sorts of things. 197 00:10:49,559 --> 00:10:53,560 Speaker 4: And so you know, the I mean the stuff they're 198 00:10:53,559 --> 00:10:56,800 Speaker 4: doing around training, you know, that's that's self self interested. 199 00:10:56,840 --> 00:10:58,400 Speaker 4: But it's great, it's awesome. It's a good thing for 200 00:10:58,520 --> 00:11:02,360 Speaker 4: NTID inc. Again, like tying all the stuff together, seven 201 00:11:02,400 --> 00:11:05,280 Speaker 4: point five billion dollar investment in New Zealand, training, one 202 00:11:05,320 --> 00:11:08,520 Speaker 4: hundred thousand people doing all this stuff, and we're so great. 203 00:11:08,600 --> 00:11:11,920 Speaker 4: I mean, just New Zealanders aren't wired that way. And 204 00:11:13,200 --> 00:11:15,080 Speaker 4: you know, I mean frankly, you know, like I know 205 00:11:15,160 --> 00:11:20,559 Speaker 4: the AS leadership here in Apja and in Seattle, and 206 00:11:21,760 --> 00:11:25,720 Speaker 4: I don't think they've had advice from you know, internally 207 00:11:25,800 --> 00:11:29,160 Speaker 4: or from their external PR people around the New Zealand 208 00:11:29,200 --> 00:11:32,840 Speaker 4: psyche and what moves the needle. And unfortunately they barked 209 00:11:32,840 --> 00:11:36,640 Speaker 4: on embarked on a PR message that went down frankly 210 00:11:36,720 --> 00:11:39,400 Speaker 4: like a cup cold sick with Kiwi's and they're paying 211 00:11:39,440 --> 00:11:40,960 Speaker 4: the price. For it now. Yeah. 212 00:11:41,040 --> 00:11:45,520 Speaker 1: Look, I had an off the record briefing with AWS 213 00:11:45,640 --> 00:11:48,760 Speaker 1: just last week, which you know I can't talk about. 214 00:11:48,800 --> 00:11:50,439 Speaker 1: I agreed to keep it off the record, but I 215 00:11:50,480 --> 00:11:53,480 Speaker 1: will say I came out of that meeting thinking just 216 00:11:53,600 --> 00:11:55,760 Speaker 1: trouble ahead here because a lot of the questions that 217 00:11:55,800 --> 00:11:59,040 Speaker 1: I've had over the years and other journalists have had, 218 00:11:59,080 --> 00:12:03,280 Speaker 1: they still weren't answering. And to go out and not 219 00:12:03,400 --> 00:12:06,040 Speaker 1: have credible answers to these questions, it was going to 220 00:12:06,559 --> 00:12:07,280 Speaker 1: go badly. 221 00:12:07,440 --> 00:12:10,320 Speaker 4: The thing is there, right, So I understand that from 222 00:12:10,400 --> 00:12:13,599 Speaker 4: a corporate strategy perspective, they don't want to tell you 223 00:12:14,360 --> 00:12:16,840 Speaker 4: the truth and the whole story. I get that, but 224 00:12:16,920 --> 00:12:19,240 Speaker 4: then don't tell a story. And so if your thing 225 00:12:19,320 --> 00:12:21,959 Speaker 4: is we're secretive, we're not going to talk about this, 226 00:12:22,640 --> 00:12:25,760 Speaker 4: then then you know, when you're briefing your PR people, 227 00:12:26,040 --> 00:12:29,319 Speaker 4: you say, right, our strategy, our corporate policy is we're 228 00:12:29,360 --> 00:12:31,640 Speaker 4: not going to disclose this, this, and this. Then the 229 00:12:31,679 --> 00:12:36,319 Speaker 4: advice should be okay. Given that lead with we're launching 230 00:12:36,320 --> 00:12:38,840 Speaker 4: the New Zealand region. This is an amazing thing, as 231 00:12:38,880 --> 00:12:42,400 Speaker 4: we'll talk about later on. Don't say anything else. Don't 232 00:12:42,440 --> 00:12:45,680 Speaker 4: say about investment, don't say about all of those other things, 233 00:12:46,120 --> 00:12:49,520 Speaker 4: and honestly, the story would have been so positive because 234 00:12:49,520 --> 00:12:53,480 Speaker 4: it's a really good thing. But whether it's arrogant, whether 235 00:12:53,640 --> 00:12:57,280 Speaker 4: it's ignorant. For some reason, they felt they had to 236 00:12:57,280 --> 00:13:00,880 Speaker 4: embark on this common strategy and it's blown up in 237 00:13:00,920 --> 00:13:01,319 Speaker 4: their face. 238 00:13:01,600 --> 00:13:04,680 Speaker 1: And then on top of that, we had the political 239 00:13:04,760 --> 00:13:08,079 Speaker 1: angle we had lux and come in and look to 240 00:13:08,840 --> 00:13:10,960 Speaker 1: be fair to the Prime minister. I don't think he 241 00:13:11,000 --> 00:13:15,840 Speaker 1: actually tried to explicitly claim credit for this. What he 242 00:13:15,880 --> 00:13:18,920 Speaker 1: did is his typical rah, rah, isn't this great? You 243 00:13:18,920 --> 00:13:21,080 Speaker 1: know anything at the moment he's trying to jump on 244 00:13:21,160 --> 00:13:25,080 Speaker 1: that shows growth of any description, and he hyped that up. 245 00:13:25,120 --> 00:13:28,679 Speaker 1: He was probably given all the key numbers from AWS 246 00:13:28,760 --> 00:13:31,480 Speaker 1: and his own team, which were exactly the key numbers 247 00:13:31,840 --> 00:13:34,880 Speaker 1: that justinto Ardern was talking about, you know, four years ago, 248 00:13:35,800 --> 00:13:38,600 Speaker 1: and he did his typical sort of you know, hype 249 00:13:38,679 --> 00:13:43,360 Speaker 1: machine shtick, and but it just it just politicized it 250 00:13:43,600 --> 00:13:47,600 Speaker 1: at a time when he's under pressure and there isn't 251 00:13:47,640 --> 00:13:50,520 Speaker 1: much good news. That really didn't help, did it, well. 252 00:13:50,360 --> 00:13:53,120 Speaker 4: I mean it didn't. And you know, like I've talked 253 00:13:53,120 --> 00:13:54,839 Speaker 4: to Chris a few times in the past. He has 254 00:13:54,840 --> 00:13:57,600 Speaker 4: my phone number, you know, like it was obvious what 255 00:13:57,679 --> 00:14:00,400 Speaker 4: was going to happen. All the editors were gonna say, like, hey, 256 00:14:00,400 --> 00:14:02,360 Speaker 4: what's the way to position this? I want to make 257 00:14:02,400 --> 00:14:05,199 Speaker 4: some political capital, I want to get some wins. How 258 00:14:05,200 --> 00:14:08,120 Speaker 4: do I position this? And there the story is, Wow, 259 00:14:08,200 --> 00:14:10,760 Speaker 4: this is a great thing. We are such a powerhouse 260 00:14:10,760 --> 00:14:14,440 Speaker 4: of technology in New Zealand that AWS is investing and 261 00:14:14,559 --> 00:14:17,160 Speaker 4: is building a region here, and that's such a win. 262 00:14:17,559 --> 00:14:21,200 Speaker 4: Technology is such an opportunity for us. Success stories like 263 00:14:21,320 --> 00:14:24,280 Speaker 4: zero and vend or whatever are now going to find 264 00:14:24,320 --> 00:14:27,480 Speaker 4: it easier to get ahead because of this, you know, 265 00:14:27,680 --> 00:14:30,840 Speaker 4: this regional buildout, and no one could argue with that 266 00:14:30,840 --> 00:14:34,400 Speaker 4: because that's actually the truth. But instead, you know, like 267 00:14:34,680 --> 00:14:37,040 Speaker 4: I don't think I mean, I don't know like the 268 00:14:38,000 --> 00:14:40,440 Speaker 4: pundits within the National Party when they're doing their polling, 269 00:14:40,520 --> 00:14:45,160 Speaker 4: do they see this as neutral or or a loss 270 00:14:45,200 --> 00:14:47,000 Speaker 4: because I don't think it's a political win. 271 00:14:47,440 --> 00:14:50,920 Speaker 1: It's definitely not. And look, you know the issue about 272 00:14:50,960 --> 00:14:54,480 Speaker 1: not building the data centers, right, it's disappointing. We want 273 00:14:54,520 --> 00:14:58,680 Speaker 1: to see more infrastructure come in, but the reality is 274 00:14:58,680 --> 00:15:02,640 Speaker 1: is there three data centers that a lot of Amazon 275 00:15:02,640 --> 00:15:05,640 Speaker 1: equipment will be going into, if not already in there. 276 00:15:06,640 --> 00:15:08,440 Speaker 1: This is not unusual around the world. I mean, there's 277 00:15:08,440 --> 00:15:11,800 Speaker 1: a big operator in Asia, Pacific air Trunk, huge data 278 00:15:11,840 --> 00:15:15,520 Speaker 1: center operator. It works with the big three Aws, Microsoft, 279 00:15:15,600 --> 00:15:19,560 Speaker 1: and Google. So they don't always build their own. They 280 00:15:19,600 --> 00:15:22,920 Speaker 1: do prefer to because they have complete control over the facilities. 281 00:15:22,960 --> 00:15:27,600 Speaker 1: But colocation is a legitimate thing in the data center industry. 282 00:15:27,640 --> 00:15:31,360 Speaker 1: They sort of blew that as well by not saying that, hey, 283 00:15:31,360 --> 00:15:34,160 Speaker 1: look we tried to build, it didn't work out, we're 284 00:15:34,160 --> 00:15:34,800 Speaker 1: co locating. 285 00:15:35,000 --> 00:15:35,240 Speaker 2: Yeah. 286 00:15:35,280 --> 00:15:38,160 Speaker 4: The difficulty here is that, and I do sort of 287 00:15:38,200 --> 00:15:43,000 Speaker 4: feel freight os here is that the hyperscalars are very 288 00:15:43,040 --> 00:15:47,080 Speaker 4: sensitive about the environmental impact of hyper scale data centers, 289 00:15:47,120 --> 00:15:51,240 Speaker 4: and so they've invested hugely in green energy, so green 290 00:15:51,360 --> 00:15:55,359 Speaker 4: energy in terms of input, but also low power servers 291 00:15:55,440 --> 00:15:58,880 Speaker 4: and really interesting ip around calling of the service because 292 00:15:58,880 --> 00:16:01,240 Speaker 4: obviously calling is a huge and when you go COLO, 293 00:16:01,880 --> 00:16:04,800 Speaker 4: you haven't got that same control. And so all of 294 00:16:04,800 --> 00:16:08,840 Speaker 4: a sudden, this strong story about you know, adel mess, 295 00:16:08,880 --> 00:16:12,560 Speaker 4: you know, green environmental credentials goes out the window. And 296 00:16:12,600 --> 00:16:15,080 Speaker 4: so it's a really tricky one. I mean, you know 297 00:16:15,160 --> 00:16:18,040 Speaker 4: the fact that they've done forward deals for green energy, 298 00:16:18,400 --> 00:16:22,880 Speaker 4: which as almost everyone in New Zealand knows actually is 299 00:16:22,880 --> 00:16:25,040 Speaker 4: irrelevant because you're taking from the grid and when the 300 00:16:25,080 --> 00:16:28,400 Speaker 4: grid hasn't got enough, you burn coal huntly. So again, 301 00:16:28,640 --> 00:16:32,560 Speaker 4: like you know, I don't know what the comms and 302 00:16:32,600 --> 00:16:35,400 Speaker 4: PR team within the organization are paid, but I can 303 00:16:35,440 --> 00:16:38,920 Speaker 4: imagine imagine it's quite a lot to not simply say, 304 00:16:38,960 --> 00:16:41,360 Speaker 4: you know, do a chat GBT, hey tell me about 305 00:16:41,400 --> 00:16:44,640 Speaker 4: the New Zealand energy sector and what happens when there's 306 00:16:44,680 --> 00:16:48,800 Speaker 4: not much power, you know, like it's not rocket science. Yeah, 307 00:16:49,200 --> 00:16:51,760 Speaker 4: they've tried. I mean, I think it's probably a situation 308 00:16:51,840 --> 00:16:55,320 Speaker 4: where you've got a big multinational is trying to have 309 00:16:55,720 --> 00:16:59,960 Speaker 4: a somewhat consistent approach in terms of regional business development, 310 00:17:00,240 --> 00:17:02,920 Speaker 4: and we see we saw the perils with that. 311 00:17:03,040 --> 00:17:06,480 Speaker 1: I guess the way AWS and Microsoft, to a lesser extent, 312 00:17:06,520 --> 00:17:09,200 Speaker 1: they have a little bit more autonomy in each country, 313 00:17:09,280 --> 00:17:13,240 Speaker 1: but Google and Aws, Apple, it's the company line everywhere. 314 00:17:13,320 --> 00:17:15,520 Speaker 1: So trying to have some nuance for the New Zealand 315 00:17:15,560 --> 00:17:17,719 Speaker 1: market isn't really going to cut it when you're talking 316 00:17:17,760 --> 00:17:22,200 Speaker 1: to Singapore or Seattle. So that is a problem. But 317 00:17:22,600 --> 00:17:25,800 Speaker 1: the energy thing did get a lot of play this week, 318 00:17:25,880 --> 00:17:29,320 Speaker 1: and I guess the fundamental issue is Kiwis are concerned 319 00:17:29,320 --> 00:17:34,200 Speaker 1: that they've heard a lot about the AI applications in 320 00:17:34,320 --> 00:17:37,879 Speaker 1: data centers. Data center is being built specifically for AI. 321 00:17:38,040 --> 00:17:42,080 Speaker 1: Zuckerberg is building one in Louisiana, fifty billion dollar data center. 322 00:17:42,840 --> 00:17:45,760 Speaker 1: The energy needs for that are massive. So they're looking 323 00:17:45,760 --> 00:17:47,560 Speaker 1: around the world and going is that going to happen 324 00:17:47,600 --> 00:17:51,080 Speaker 1: to us? We're already seeing mills shut down and potential 325 00:17:51,119 --> 00:17:54,960 Speaker 1: brownouts when you know the lake levels fall. Is this 326 00:17:55,040 --> 00:17:59,760 Speaker 1: going to crowd out demand for existing industry and for consumers? 327 00:17:59,840 --> 00:18:02,000 Speaker 1: Is well? Is that an issue when you look at 328 00:18:02,080 --> 00:18:05,320 Speaker 1: the evolution of the cloud industry that you've been really 329 00:18:05,359 --> 00:18:08,280 Speaker 1: close to for a long time, that is signals real 330 00:18:08,320 --> 00:18:09,159 Speaker 1: trouble ahead for it. 331 00:18:09,280 --> 00:18:11,639 Speaker 4: Yeah, it is. And I mean, you know, like both 332 00:18:12,040 --> 00:18:14,800 Speaker 4: Microsoft Nata are spending a lot of time shouting from 333 00:18:14,800 --> 00:18:17,879 Speaker 4: the rooftops about all their green energy investments, They're not 334 00:18:17,920 --> 00:18:20,320 Speaker 4: shouting quite so high from the rooftops about the number 335 00:18:20,359 --> 00:18:23,080 Speaker 4: of nuclear plants that are being built near their data centers. 336 00:18:23,119 --> 00:18:25,959 Speaker 4: I mean, I think this is a broader ends in conversation. 337 00:18:27,200 --> 00:18:30,920 Speaker 4: And you know, when the whole t Y situation happened, 338 00:18:31,040 --> 00:18:34,639 Speaker 4: and you know whether or not they were going to 339 00:18:34,680 --> 00:18:38,040 Speaker 4: get a good deal on their energy. We need a 340 00:18:38,080 --> 00:18:41,200 Speaker 4: conversation about what our priorities are. We need a conversation 341 00:18:41,280 --> 00:18:45,800 Speaker 4: about resilience. I mean generally speaking, generation happens in the 342 00:18:46,200 --> 00:18:49,720 Speaker 4: South Island, the load is in the North Island. Note 343 00:18:49,840 --> 00:18:52,840 Speaker 4: that data Grid is building is hoping to build a 344 00:18:53,920 --> 00:18:57,639 Speaker 4: large scale data center close down to Manipuri, so that 345 00:18:57,680 --> 00:19:01,200 Speaker 4: will be will be great, will's actually low and generation 346 00:19:01,280 --> 00:19:05,760 Speaker 4: capacity close by. But the conversation for us as New 347 00:19:05,800 --> 00:19:09,200 Speaker 4: Zealanders is do we want aluminium smelting to happen here? 348 00:19:09,280 --> 00:19:11,520 Speaker 4: Do we want steal smelting to happen to happen here? 349 00:19:11,560 --> 00:19:14,920 Speaker 4: Do we want pop and paper mills? And it's way 350 00:19:15,000 --> 00:19:17,400 Speaker 4: beyond the topic of this conversation, but I think that 351 00:19:18,400 --> 00:19:21,359 Speaker 4: no politician is given ever going to ask us those 352 00:19:21,480 --> 00:19:26,320 Speaker 4: questions because they want to put forward a win win scenario. 353 00:19:26,520 --> 00:19:29,359 Speaker 4: The reality is there is no win win scenario, and 354 00:19:30,440 --> 00:19:33,960 Speaker 4: either we need more generation or we need to accept 355 00:19:34,000 --> 00:19:36,080 Speaker 4: the nuclear is going to happen, or build some dams, 356 00:19:36,160 --> 00:19:38,440 Speaker 4: or we're going to do sola or wind, or we're 357 00:19:38,440 --> 00:19:40,800 Speaker 4: not going to have energy to use, or we're not 358 00:19:40,840 --> 00:19:44,080 Speaker 4: going to have large scale industrial users of energy. And 359 00:19:44,240 --> 00:19:45,960 Speaker 4: these are all leavers that we can pull, but we 360 00:19:46,000 --> 00:19:48,119 Speaker 4: can't pull all the leavers at once and expect the 361 00:19:48,119 --> 00:19:51,160 Speaker 4: stuff to work. And I sort of bemoan the fact 362 00:19:51,160 --> 00:19:54,679 Speaker 4: in New Zealand that the discourse tends to be quite 363 00:19:54,800 --> 00:19:58,320 Speaker 4: immature in terms of understanding those trade off aspects. 364 00:19:58,520 --> 00:20:01,119 Speaker 1: And look, you know, the likes of America and Contact 365 00:20:01,359 --> 00:20:06,840 Speaker 1: Energy contacted to deal with Microsoft to take geothermal capacity. 366 00:20:07,680 --> 00:20:09,960 Speaker 1: You know, they point out, look for us to invest 367 00:20:10,119 --> 00:20:13,959 Speaker 1: in additional capacity, we need sort of anchor tenance, just 368 00:20:14,000 --> 00:20:15,880 Speaker 1: like you need an anchor tenant in a data center 369 00:20:15,920 --> 00:20:19,880 Speaker 1: if you're leasing that space out, you need commitments from 370 00:20:19,920 --> 00:20:23,040 Speaker 1: a couple of big companies that they're going to take 371 00:20:23,200 --> 00:20:28,040 Speaker 1: and pay for over the coming decades your output. So 372 00:20:29,080 --> 00:20:31,040 Speaker 1: how many companies are putting up their hand to do that. 373 00:20:31,520 --> 00:20:33,840 Speaker 1: The reality is the growth is in the digital economy. 374 00:20:33,960 --> 00:20:37,280 Speaker 1: So in that sense, you know, it's only natural that 375 00:20:37,480 --> 00:20:39,480 Speaker 1: the data center companies are the ones that are sort 376 00:20:39,520 --> 00:20:40,119 Speaker 1: of anchoring this. 377 00:20:40,400 --> 00:20:43,119 Speaker 4: Yeah, that's right in I mean, you know, like I 378 00:20:43,160 --> 00:20:45,560 Speaker 4: know as you do, Microsoft has actually built out so 379 00:20:45,680 --> 00:20:48,359 Speaker 4: and data center in terms of ata WS, I don't 380 00:20:48,400 --> 00:20:51,280 Speaker 4: actually understand, like I understand that when you're at that 381 00:20:51,359 --> 00:20:56,119 Speaker 4: scale you can actually buy directly from the gentailors, and 382 00:20:56,200 --> 00:20:59,240 Speaker 4: so you know it's coming from the grid, but if 383 00:20:59,240 --> 00:21:01,880 Speaker 4: you squint, it is it is sort of directly from 384 00:21:01,880 --> 00:21:05,040 Speaker 4: the generators in the a US case because they're COLO. 385 00:21:05,560 --> 00:21:09,399 Speaker 4: I'm not sure how that actually works. I mean, yes, 386 00:21:09,520 --> 00:21:12,480 Speaker 4: there will be more energy consumed because AS has a 387 00:21:12,520 --> 00:21:15,520 Speaker 4: region here, but I'm not sure you can then conflate 388 00:21:15,560 --> 00:21:19,679 Speaker 4: that to ADS having a direct connection or direct direct 389 00:21:19,720 --> 00:21:27,200 Speaker 4: commercial relationship with a generator, given that they are buying capacity. Eventually, 390 00:21:27,520 --> 00:21:31,240 Speaker 4: essentially they are leasing space in a pre existing data center. 391 00:21:31,359 --> 00:21:35,920 Speaker 4: So yeah, yeah, truth truth, truth, statistics and damn liners 392 00:21:35,960 --> 00:21:36,840 Speaker 4: and all that good stuff. 393 00:21:37,200 --> 00:21:40,119 Speaker 1: Yeah. I put that question to them on Monday. Are 394 00:21:40,160 --> 00:21:42,080 Speaker 1: they going to use carbon credit? So how is this 395 00:21:42,119 --> 00:21:45,320 Speaker 1: going to work, particularly if there is real constraints on 396 00:21:45,760 --> 00:21:49,840 Speaker 1: renewable energy. Didn't get an answer on that, But look, Ben, 397 00:21:49,880 --> 00:21:52,040 Speaker 1: what are we going to get here? So presumably they've 398 00:21:52,080 --> 00:21:56,040 Speaker 1: already loaded up those data centers with Amazon equipment. Amazon 399 00:21:56,119 --> 00:22:01,280 Speaker 1: makes its own really high performance silicon chips. They'll have 400 00:22:01,560 --> 00:22:05,440 Speaker 1: probably Nvidia and other stuff maybe in coming to those 401 00:22:05,520 --> 00:22:08,200 Speaker 1: data centers. So there's a lot of equipment in there now. 402 00:22:08,520 --> 00:22:11,359 Speaker 1: They have a big roster of New Zealand clients, most 403 00:22:11,359 --> 00:22:13,560 Speaker 1: of them will be hosted at the moment in Sydney 404 00:22:14,520 --> 00:22:18,320 Speaker 1: or further afield. So do you think there will be 405 00:22:18,800 --> 00:22:23,320 Speaker 1: a big influx of applications and data into those data 406 00:22:23,320 --> 00:22:25,600 Speaker 1: centers in the coming months. I guess that that is 407 00:22:25,600 --> 00:22:28,200 Speaker 1: the good story where you start talking about lower latency 408 00:22:28,200 --> 00:22:31,879 Speaker 1: because you're closer to the data center. Know what are 409 00:22:31,920 --> 00:22:33,680 Speaker 1: some of the other benefits do you think that those 410 00:22:33,720 --> 00:22:36,280 Speaker 1: clients will will now take advantage of? 411 00:22:36,720 --> 00:22:36,920 Speaker 3: Yeah? 412 00:22:36,960 --> 00:22:39,879 Speaker 4: I mean, I guess there are three reasons why you 413 00:22:39,960 --> 00:22:46,600 Speaker 4: choose where to cite your cloud workload. Latency, data sovereignty, 414 00:22:47,119 --> 00:22:51,840 Speaker 4: and then cost and so you know, like I don't know, 415 00:22:51,920 --> 00:22:55,639 Speaker 4: I haven't looked into ad esus pricing in New Zealand 416 00:22:55,720 --> 00:22:59,440 Speaker 4: versus in Sydney. I would imagine it's the same. I'd 417 00:22:59,440 --> 00:23:04,360 Speaker 4: spect latency. You know, it's it's a long time to swim, 418 00:23:04,400 --> 00:23:07,000 Speaker 4: but it's a pretty short time for the speed of 419 00:23:07,080 --> 00:23:11,119 Speaker 4: light between Sydney and New Zealand. So for sure, some 420 00:23:11,240 --> 00:23:15,560 Speaker 4: high frequency trading workloads, potentially the latency thing is going 421 00:23:15,600 --> 00:23:19,280 Speaker 4: to be attractive, But to be honest, I discount that 422 00:23:19,520 --> 00:23:22,679 Speaker 4: a little bit. So really that the core proposition is 423 00:23:22,720 --> 00:23:26,920 Speaker 4: the data sovereignty one, and in a world where geopolitics 424 00:23:26,960 --> 00:23:30,840 Speaker 4: are becoming ever more fraught, that is even more compelling. 425 00:23:30,880 --> 00:23:33,520 Speaker 4: I mean, Australia is a very good friend of New Zealand. 426 00:23:34,280 --> 00:23:37,560 Speaker 4: But still, however, I would say, you know, I have 427 00:23:37,720 --> 00:23:40,800 Speaker 4: Don Christie ringing in my ear. Don Christy from the Catalyst, 428 00:23:40,800 --> 00:23:43,879 Speaker 4: and if he was here, he would say, as at 429 00:23:43,920 --> 00:23:47,760 Speaker 4: the end of the day is a US corporation. If 430 00:23:47,760 --> 00:23:52,359 Speaker 4: we are concerned about geopolitics, then jumping into bed with 431 00:23:52,520 --> 00:23:56,000 Speaker 4: the US corporation, notwithstanding that it has some presence here 432 00:23:56,040 --> 00:23:58,880 Speaker 4: in New Zealand, isn't a good thing. He would probably 433 00:23:58,880 --> 00:24:02,639 Speaker 4: back that up by saying, of that four hundred and 434 00:24:02,640 --> 00:24:06,680 Speaker 4: whatever million dollars with the revenue that AWS received here, 435 00:24:07,240 --> 00:24:10,840 Speaker 4: how much do they pay tax on? So all of 436 00:24:10,880 --> 00:24:14,359 Speaker 4: those questions are valid, I guess, and I have massive 437 00:24:14,359 --> 00:24:17,199 Speaker 4: amounts of respect for Don. I guess. My response to 438 00:24:17,280 --> 00:24:20,760 Speaker 4: him is the horse has already bolted. You know, the 439 00:24:20,840 --> 00:24:26,760 Speaker 4: hyperscalers Google, Microsoft, As and Ali Baba are quite just that. 440 00:24:26,880 --> 00:24:32,399 Speaker 4: They're hyperscalers. They are the ones to beat. Frankly, you 441 00:24:32,480 --> 00:24:35,679 Speaker 4: can't beat them. You're playing around the edges. And what 442 00:24:35,760 --> 00:24:38,200 Speaker 4: Ketalyst Cloud does is really really valuable, and there's a 443 00:24:38,200 --> 00:24:42,200 Speaker 4: strong value proposition. But it's a niche player, and I 444 00:24:42,280 --> 00:24:44,720 Speaker 4: think will only be a niche player. So it's a 445 00:24:44,760 --> 00:24:47,560 Speaker 4: great thing that AWS is here. There will be a 446 00:24:47,600 --> 00:24:51,600 Speaker 4: bunch of organizations that are quite excited by it. Frankly, 447 00:24:51,720 --> 00:24:56,080 Speaker 4: if I ran AWS, I would absolutely go down a 448 00:24:56,160 --> 00:24:59,280 Speaker 4: road of co location, because I'm not convinced that the 449 00:24:59,520 --> 00:25:01,320 Speaker 4: scale or of the workloads there are going to be 450 00:25:01,440 --> 00:25:05,560 Speaker 4: run in the New Zealand region as significant enough to 451 00:25:05,760 --> 00:25:09,160 Speaker 4: justify actually building the head infrastructure here. So I think 452 00:25:09,160 --> 00:25:12,760 Speaker 4: it's the right strategy. It's just been badly articulated. 453 00:25:12,920 --> 00:25:16,479 Speaker 1: Yeah. Well, maybe they came to that conclusion sort of 454 00:25:16,520 --> 00:25:20,800 Speaker 1: as we came out of the boom of the pandemic 455 00:25:20,880 --> 00:25:24,840 Speaker 1: area where there was a lot of cheap cash around 456 00:25:25,040 --> 00:25:28,560 Speaker 1: and frankly a massive surge in the digital economy as 457 00:25:28,600 --> 00:25:31,880 Speaker 1: people moved online. Maybe the thinking was in twenty twenty one, 458 00:25:32,680 --> 00:25:34,600 Speaker 1: this is going to go exponential, where it's going to 459 00:25:34,600 --> 00:25:36,600 Speaker 1: get more and more business in this part of the world, 460 00:25:36,600 --> 00:25:39,560 Speaker 1: and it didn't turn out that way. So difficulties with 461 00:25:39,640 --> 00:25:43,760 Speaker 1: the site in Auckland as side. Something obviously changed along 462 00:25:43,800 --> 00:25:44,280 Speaker 1: the line. 463 00:25:44,400 --> 00:25:47,359 Speaker 4: Yeah, And the thing is that if you're spending or 464 00:25:47,760 --> 00:25:50,880 Speaker 4: let's believe the number spending seven point five billion dollars. 465 00:25:51,560 --> 00:25:55,359 Speaker 4: Consenting issues are not a barrier. You can afford really 466 00:25:55,359 --> 00:26:00,040 Speaker 4: good lawyers. You can pay for drainage or whatever. So 467 00:26:00,119 --> 00:26:03,000 Speaker 4: I actually don't think that it was a you know, 468 00:26:03,600 --> 00:26:07,199 Speaker 4: an issue with the consenting regulatory framework in New Zealand. 469 00:26:07,520 --> 00:26:10,280 Speaker 4: I think frankly that said a lot. Actually the prize 470 00:26:10,280 --> 00:26:12,840 Speaker 4: isn't that big. Let's go co low and we can 471 00:26:12,840 --> 00:26:14,119 Speaker 4: always change if we need to. 472 00:26:14,359 --> 00:26:17,439 Speaker 1: And just finally being you know, they're talking obviously you 473 00:26:17,440 --> 00:26:21,960 Speaker 1: can put your apps and shift all your data there now, 474 00:26:22,280 --> 00:26:28,200 Speaker 1: but they talked on Monday about the AI tools such 475 00:26:28,240 --> 00:26:31,280 Speaker 1: as the Bedrock platform. This is where you can choose 476 00:26:31,359 --> 00:26:35,359 Speaker 1: what large language model you want to use. Sage Maker 477 00:26:35,440 --> 00:26:38,920 Speaker 1: their AI sort of tool sets. So that stuff is coming. 478 00:26:39,400 --> 00:26:41,800 Speaker 1: They need dedicated hardware to run that. That will be 479 00:26:41,840 --> 00:26:44,680 Speaker 1: switched on, they said later this year. We're running out 480 00:26:44,680 --> 00:26:47,320 Speaker 1: a year, so that won't be far away. Is that 481 00:26:47,359 --> 00:26:49,199 Speaker 1: an exciting thing to you? I mean, that's what this 482 00:26:49,320 --> 00:26:53,160 Speaker 1: is all about, is not simply just putting your applications 483 00:26:53,200 --> 00:26:55,879 Speaker 1: in the cloud and the efficiencies you get from that, 484 00:26:55,960 --> 00:27:01,000 Speaker 1: but actually transforming your business using IF intelligence. Is that 485 00:27:01,119 --> 00:27:02,680 Speaker 1: genuinely exciting to you? 486 00:27:03,119 --> 00:27:03,359 Speaker 2: Yeah? 487 00:27:03,359 --> 00:27:07,600 Speaker 4: That's right. So so for some context there, without being 488 00:27:07,840 --> 00:27:11,720 Speaker 4: being too geeky, you know your general sort of computer 489 00:27:11,840 --> 00:27:15,000 Speaker 4: or storage basically taking the server that was in your 490 00:27:15,000 --> 00:27:18,480 Speaker 4: cupboard and moving that into the cloud. That's pretty generic 491 00:27:18,560 --> 00:27:22,760 Speaker 4: commodity hardware that is easy to come by that has 492 00:27:22,800 --> 00:27:25,920 Speaker 4: done at scale. When you're do an AI workloads, that's 493 00:27:25,960 --> 00:27:29,720 Speaker 4: done generally on what's called GPU's graphical processing units that 494 00:27:30,040 --> 00:27:32,480 Speaker 4: come from the likes of n Video. But as you 495 00:27:32,520 --> 00:27:35,960 Speaker 4: pointed out, a WS has its own trainingum silicon and 496 00:27:36,000 --> 00:27:38,040 Speaker 4: that stuff is a little bit harder. The supply chain 497 00:27:38,119 --> 00:27:40,600 Speaker 4: for that's a little bit longer, to be honest. The 498 00:27:40,640 --> 00:27:44,880 Speaker 4: other reason is that everyone uses computing storage, not everyone 499 00:27:45,080 --> 00:27:50,199 Speaker 4: is training large language models and needs GPUs. So I 500 00:27:50,200 --> 00:27:52,320 Speaker 4: think there's there's there's a couple of answers to that. 501 00:27:53,880 --> 00:27:58,040 Speaker 4: Generally speaking, training l l MS is something that isn't 502 00:27:58,520 --> 00:28:04,720 Speaker 4: particularly critical in terms of latency or even location. So 503 00:28:05,440 --> 00:28:09,480 Speaker 4: if I was running a SaaS company and latency was 504 00:28:09,520 --> 00:28:12,320 Speaker 4: a huge issue, and I was also doing some AI stuff, 505 00:28:12,520 --> 00:28:16,040 Speaker 4: I would probably strategically say putting data sovereignty aside for 506 00:28:16,080 --> 00:28:20,000 Speaker 4: a moment, I'm going to train my llms wherever it's 507 00:28:20,119 --> 00:28:25,359 Speaker 4: cheapest and easiest and most accessible, but the actual application 508 00:28:25,480 --> 00:28:28,320 Speaker 4: I'm going to run as close to the demand as possible. 509 00:28:28,960 --> 00:28:32,280 Speaker 4: So I guess the question that comes out of it is, 510 00:28:32,640 --> 00:28:38,280 Speaker 4: if you're doing AI type workloads, how concerned are you 511 00:28:38,560 --> 00:28:42,480 Speaker 4: about whether those happen here or elsewhere? Given that latency 512 00:28:42,600 --> 00:28:45,320 Speaker 4: generally isn't an issue. Again, it comes back to the 513 00:28:45,360 --> 00:28:48,360 Speaker 4: sovereignty thing. The other thing I would say, which is 514 00:28:48,400 --> 00:28:51,200 Speaker 4: really interesting, and it's going to be a bit of 515 00:28:51,200 --> 00:28:54,239 Speaker 4: a voyage on gequery, sorry, but there's a lot of 516 00:28:54,280 --> 00:28:58,440 Speaker 4: stuff happening around AI at the edge. So basically, you 517 00:28:59,240 --> 00:29:02,600 Speaker 4: chuck a bunch of data into a big hyperscale data 518 00:29:02,640 --> 00:29:06,360 Speaker 4: center in order to train a model, but then you 519 00:29:06,400 --> 00:29:08,600 Speaker 4: deploy that model at the edge so that you know, 520 00:29:08,680 --> 00:29:13,920 Speaker 4: let's say, for example, you're doing a connected camera for 521 00:29:14,000 --> 00:29:16,360 Speaker 4: home security, right, and you want to be able to 522 00:29:16,400 --> 00:29:20,640 Speaker 4: identify when someone comes, you know, different types of behavior 523 00:29:21,320 --> 00:29:23,880 Speaker 4: sentiment analysis, look at their face, see if they're happy 524 00:29:23,920 --> 00:29:27,960 Speaker 4: or sad or aggressive or whatever. Probably the approach is 525 00:29:28,120 --> 00:29:34,360 Speaker 4: you're going to train a bunch of models to understand happy, sad, aggressive, whatever, 526 00:29:35,360 --> 00:29:38,000 Speaker 4: and then you're going to get a really small, relatively 527 00:29:38,040 --> 00:29:41,680 Speaker 4: small model, and you're going to implement that on your camera, 528 00:29:42,160 --> 00:29:44,440 Speaker 4: so you don't need to send data back to some 529 00:29:44,520 --> 00:29:49,520 Speaker 4: data center wherever to say, oh, hey, something's happening, tell 530 00:29:49,560 --> 00:29:51,600 Speaker 4: me what it is. You'll do that at the edge. 531 00:29:51,600 --> 00:29:55,400 Speaker 4: So it's really interesting what's happening around distributed cloud distributed 532 00:29:55,520 --> 00:29:59,880 Speaker 4: data processing. And the interesting thing with the hyperscale is 533 00:29:59,880 --> 00:30:03,040 Speaker 4: is they are very much sort of predicated on a 534 00:30:03,120 --> 00:30:07,560 Speaker 4: quite a centralized model, but that's becoming a lot more granular, 535 00:30:07,760 --> 00:30:11,040 Speaker 4: and all of the hyper scalers are offering also sort 536 00:30:11,040 --> 00:30:14,120 Speaker 4: of models at the edge or you know, compute at 537 00:30:14,120 --> 00:30:17,240 Speaker 4: the edge or l l ms at the edge. But 538 00:30:17,320 --> 00:30:21,040 Speaker 4: it's really interesting in terms of this bifurcation around mass 539 00:30:21,120 --> 00:30:26,640 Speaker 4: processing at scale in a centralized manner and edge processing 540 00:30:27,760 --> 00:30:31,320 Speaker 4: where you need it, where there's zero latency and you're 541 00:30:31,320 --> 00:30:33,520 Speaker 4: not sending a bunch of data backwards and forward. 542 00:30:33,560 --> 00:30:35,280 Speaker 1: But just finally, Ben, as you know, how big a 543 00:30:35,360 --> 00:30:38,280 Speaker 1: blow is this for AWS? I mean for the likes 544 00:30:38,320 --> 00:30:41,920 Speaker 1: of Kiwi Bank and those who are buying capacity in 545 00:30:41,960 --> 00:30:44,480 Speaker 1: that data center. You know, they'll probably just shrug and go, well, 546 00:30:44,520 --> 00:30:47,640 Speaker 1: that was a bad week for AWS, but they're still 547 00:30:47,840 --> 00:30:51,160 Speaker 1: providing great services to us. Do you think it damages 548 00:30:51,200 --> 00:30:54,440 Speaker 1: the company's credibility in any sustainable way. 549 00:30:54,600 --> 00:30:57,560 Speaker 4: It's a really great question because you know, people like 550 00:30:57,600 --> 00:31:00,000 Speaker 4: you and I the tech sector are spending a huge, 551 00:31:00,080 --> 00:31:03,720 Speaker 4: huge amount of cycles sort of second guessing what as 552 00:31:03,800 --> 00:31:07,320 Speaker 4: did and didn't do and should have done. The reality 553 00:31:07,480 --> 00:31:12,040 Speaker 4: is they're very smart individuals. They've built a massive, massive 554 00:31:12,080 --> 00:31:15,200 Speaker 4: business and in a year's time isn't going to make 555 00:31:15,280 --> 00:31:19,000 Speaker 4: any difference to the growth they have seen or would 556 00:31:19,000 --> 00:31:22,480 Speaker 4: have seen in New Zealand. Absolutely not. So it's absolutely 557 00:31:22,520 --> 00:31:24,880 Speaker 4: a mistake and it's kind of an amateur screw up, 558 00:31:25,240 --> 00:31:27,640 Speaker 4: but it really really doesn't matter from the scheme of things, 559 00:31:28,080 --> 00:31:32,360 Speaker 4: Probably that the biggest damage, to be fair is political damage. 560 00:31:32,400 --> 00:31:36,280 Speaker 4: As I say, I think Prime Minister should have got 561 00:31:36,360 --> 00:31:40,000 Speaker 4: better advice, whereas advisors should have sought better advice. You know, 562 00:31:40,160 --> 00:31:42,240 Speaker 4: like again, that impact is not going to be you know, 563 00:31:42,320 --> 00:31:44,000 Speaker 4: the election isn't going to be one or lost on 564 00:31:45,280 --> 00:31:49,040 Speaker 4: a gas around an a OS launch. But yeah, I 565 00:31:49,080 --> 00:31:49,760 Speaker 4: don't think it's. 566 00:31:49,600 --> 00:31:52,920 Speaker 1: Helped thanks to being keeps for his analysis of aws's 567 00:31:53,080 --> 00:31:57,040 Speaker 1: bumpy launch, He's right, a WS will happily roll on 568 00:31:57,600 --> 00:32:01,440 Speaker 1: picking up business for its co located data centers and 569 00:32:01,560 --> 00:32:04,600 Speaker 1: power to them. There's a lot of great innovation that's 570 00:32:04,640 --> 00:32:07,120 Speaker 1: going to be available right on our doorstep. But there's 571 00:32:07,120 --> 00:32:11,160 Speaker 1: also a lesson for big tech. People are concerned about 572 00:32:11,200 --> 00:32:14,720 Speaker 1: the data center boom, what it means for energy prices, 573 00:32:15,520 --> 00:32:19,800 Speaker 1: for data sovereignty, for our ability to steer our own destiny. 574 00:32:19,880 --> 00:32:23,520 Speaker 1: As a tiny handful of companies become central to everything 575 00:32:23,600 --> 00:32:28,400 Speaker 1: we do in the digital world, people have questions. Trust 576 00:32:28,560 --> 00:32:31,400 Speaker 1: is fragile, and it definitely took a dent this week. 577 00:32:31,520 --> 00:32:33,920 Speaker 1: Hope you enjoyed this special episode. Let me know what 578 00:32:33,960 --> 00:32:37,640 Speaker 1: you think about the AWS Data Center region launch back 579 00:32:37,640 --> 00:32:39,840 Speaker 1: to normal next week. I'll catch you next Thursday with 580 00:32:39,920 --> 00:32:42,160 Speaker 1: another episode of the Business of Tech.