1 00:00:09,093 --> 00:00:12,693 Speaker 1: You're listening to a podcast from News Talk ZEDB. Follow 2 00:00:12,773 --> 00:00:16,173 Speaker 1: this and our wide range of podcasts now on iHeartRadio. 3 00:00:16,613 --> 00:00:19,453 Speaker 2: What do we make of the Amazon Web services announcements 4 00:00:19,493 --> 00:00:22,133 Speaker 2: and the data centers? Peter Griffin is a Wellington based 5 00:00:22,173 --> 00:00:24,893 Speaker 2: science and technology journalist with twenty years experience in the 6 00:00:24,893 --> 00:00:29,253 Speaker 2: New Zealand media covering science, technology and media and he 7 00:00:29,373 --> 00:00:31,893 Speaker 2: joins us on the line now and he's been all 8 00:00:31,933 --> 00:00:33,812 Speaker 2: over the story. Peter, A very good afternoon to you. 9 00:00:34,453 --> 00:00:37,772 Speaker 3: Good afternoon, Hey, let's go right back to basics. A. 10 00:00:38,533 --> 00:00:41,533 Speaker 3: What are we talking about here? What is a data center? 11 00:00:41,573 --> 00:00:44,092 Speaker 3: And what are they looking at building aws? 12 00:00:44,693 --> 00:00:48,373 Speaker 4: Well, you know, we've all moved into the digital world 13 00:00:48,412 --> 00:00:50,732 Speaker 4: over the last twenty years, and it used to be 14 00:00:50,772 --> 00:00:54,093 Speaker 4: that a lot of those computer service they sat all 15 00:00:54,133 --> 00:00:57,933 Speaker 4: over the place and businesses all over our major cities. 16 00:00:57,933 --> 00:01:00,853 Speaker 4: And it's become a lot more efficient over the last 17 00:01:00,973 --> 00:01:04,212 Speaker 4: ten to fifteen years to move all of that computer 18 00:01:04,293 --> 00:01:09,212 Speaker 4: equipment into centralized facilities called data centers, where there are 19 00:01:09,333 --> 00:01:12,053 Speaker 4: racks and racks and racks of this equipment. It's more 20 00:01:12,053 --> 00:01:17,333 Speaker 4: efficient to power these facilities, easier to secure them, and 21 00:01:17,413 --> 00:01:19,533 Speaker 4: to have lots of computer capacity in one place. So 22 00:01:19,572 --> 00:01:22,813 Speaker 4: that's not a new thing, but we've sort of missed 23 00:01:22,853 --> 00:01:26,573 Speaker 4: out until recently on the big data center boom that's 24 00:01:26,572 --> 00:01:31,053 Speaker 4: happened overseas, where the likes of Amazon, Microsoft, and Google 25 00:01:31,093 --> 00:01:34,533 Speaker 4: have been spending billions and billions on these data centers. 26 00:01:35,093 --> 00:01:38,253 Speaker 4: About four years ago, Amazon said it was actually going 27 00:01:38,292 --> 00:01:41,373 Speaker 4: to start building these data centers in New Zealand. It's 28 00:01:41,453 --> 00:01:44,213 Speaker 4: been a long process since then. The economy has sort 29 00:01:44,212 --> 00:01:47,093 Speaker 4: of tanked since they made this announcement. But just today 30 00:01:47,133 --> 00:01:50,493 Speaker 4: they announced they're open for business finally for these data 31 00:01:50,533 --> 00:01:54,133 Speaker 4: centers in the Auckland region and the investment they claim 32 00:01:54,213 --> 00:01:57,293 Speaker 4: will be seven and a half billion over fifteen years 33 00:01:57,333 --> 00:02:00,813 Speaker 4: between building the things, staffing them, and the training that 34 00:02:01,333 --> 00:02:04,253 Speaker 4: goes along with encouraging kiwis to use them. 35 00:02:05,053 --> 00:02:09,093 Speaker 3: So they are actually going to build the centers, and 36 00:02:09,293 --> 00:02:12,893 Speaker 3: there's talk about three availability zones, so they have committed 37 00:02:12,893 --> 00:02:17,693 Speaker 3: to actually building these three physical structures in New Zealand. 38 00:02:18,493 --> 00:02:22,213 Speaker 4: Well it's interesting because they I think at one point 39 00:02:22,333 --> 00:02:25,973 Speaker 4: they set out to physically build at least one data center, 40 00:02:26,013 --> 00:02:28,813 Speaker 4: but they ran into trouble in West Auckland with the 41 00:02:28,853 --> 00:02:32,453 Speaker 4: site that they chose with drainage there, so all works 42 00:02:32,493 --> 00:02:35,613 Speaker 4: stopped on that and it hasn't continued. What they've pivoted 43 00:02:35,693 --> 00:02:40,733 Speaker 4: to is basically using co location, which is putting all 44 00:02:40,733 --> 00:02:44,453 Speaker 4: of your equipment and infrastructure into other people's data centers. Now, 45 00:02:44,453 --> 00:02:48,093 Speaker 4: they won't say who those companies are. I've got my 46 00:02:48,133 --> 00:02:51,053 Speaker 4: suspicions because there's only a small number of companies in 47 00:02:51,093 --> 00:02:55,853 Speaker 4: the Auckland region doing it. Aws Amazon say, it doesn't 48 00:02:55,893 --> 00:02:58,493 Speaker 4: really matter who it is. We can tell you from 49 00:02:58,493 --> 00:03:01,093 Speaker 4: a security point of view, but we have three facilities 50 00:03:01,613 --> 00:03:04,252 Speaker 4: that if there's an emergency, like one of them is 51 00:03:04,293 --> 00:03:08,693 Speaker 4: knocked out, there's always going to be backup capacity. They 52 00:03:08,733 --> 00:03:12,373 Speaker 4: will automatically switch over to these other two that are 53 00:03:12,413 --> 00:03:14,933 Speaker 4: within one hundred kilometers of the other one. 54 00:03:15,333 --> 00:03:18,693 Speaker 3: So is that just going to be a temporary situation 55 00:03:19,013 --> 00:03:22,412 Speaker 3: cookooing and on those other centers and they're eventually going 56 00:03:22,453 --> 00:03:25,773 Speaker 3: to continue to build they actually build these things because 57 00:03:25,972 --> 00:03:29,252 Speaker 3: I say that that drainage issue has been solved now 58 00:03:30,213 --> 00:03:31,413 Speaker 3: or is that just going to be it? Are they 59 00:03:31,493 --> 00:03:33,773 Speaker 3: just are they going to slide them in with existing 60 00:03:33,853 --> 00:03:34,493 Speaker 3: data centers? 61 00:03:34,973 --> 00:03:37,333 Speaker 4: Well, for the moment that they're saying they've got the 62 00:03:37,373 --> 00:03:40,973 Speaker 4: three data centers they're co locating at the moment, But look, 63 00:03:41,013 --> 00:03:44,853 Speaker 4: you know, this is a very fast growing area, particularly 64 00:03:44,853 --> 00:03:48,333 Speaker 4: with the advent of artificial intelligence, the likes of chat, 65 00:03:48,413 --> 00:03:52,013 Speaker 4: GPT type services. Amazon service up its own services and 66 00:03:52,093 --> 00:03:55,173 Speaker 4: allows any New Zealand business to do that. So the 67 00:03:55,253 --> 00:03:59,133 Speaker 4: growth is pretty much exponential in the data and the 68 00:03:59,133 --> 00:04:01,453 Speaker 4: computing power that you need to run all of this 69 00:04:02,053 --> 00:04:05,453 Speaker 4: artificial intelligence stuff. So there's a good chance they do 70 00:04:05,573 --> 00:04:08,933 Speaker 4: own that patch of land. If they can remediate that proper, 71 00:04:09,333 --> 00:04:14,093 Speaker 4: properly in the coming years, they will have space to 72 00:04:14,373 --> 00:04:18,053 Speaker 4: build another one, and I think we'll see Microsoft, Google 73 00:04:18,093 --> 00:04:19,373 Speaker 4: and others doing that as well. 74 00:04:20,053 --> 00:04:23,093 Speaker 3: So the seven point five billion dollar figure that we've 75 00:04:23,133 --> 00:04:26,413 Speaker 3: been hearing in the one thousand jobs, is that still 76 00:04:26,453 --> 00:04:28,733 Speaker 3: a thing. Is that still going to happen even if 77 00:04:28,733 --> 00:04:31,613 Speaker 3: there isn't just yet any construction. 78 00:04:33,013 --> 00:04:37,773 Speaker 4: Yeah, well, they've committed again today to all of those figures. 79 00:04:37,853 --> 00:04:40,533 Speaker 4: So the thousand staff, they've been pretty clear from the 80 00:04:40,573 --> 00:04:44,093 Speaker 4: start that most of those staff will not be in 81 00:04:44,213 --> 00:04:51,493 Speaker 4: construction or in aws own workforce. The thousands, the thousand 82 00:04:52,173 --> 00:04:56,213 Speaker 4: jobs that are going to be created come downstream, so 83 00:04:56,333 --> 00:05:00,333 Speaker 4: all the startups that it's easier for them to spin 84 00:05:00,453 --> 00:05:03,333 Speaker 4: up a company and use all of these AI tools 85 00:05:03,893 --> 00:05:07,853 Speaker 4: in the cloud in this local data center. They're saying 86 00:05:07,853 --> 00:05:09,893 Speaker 4: that there's going to be more economic development as a 87 00:05:09,933 --> 00:05:12,573 Speaker 4: result of more companies using the cloud. So that's where 88 00:05:12,613 --> 00:05:15,973 Speaker 4: the bulk of the job numbers come from. One hundred 89 00:05:16,013 --> 00:05:19,293 Speaker 4: thousand ki We's trained sounds really impressive. They've already done 90 00:05:19,333 --> 00:05:22,413 Speaker 4: fifty thousand, but it's hard to know exactly. You know, 91 00:05:22,413 --> 00:05:24,613 Speaker 4: it's that a free sort of two hour course you 92 00:05:24,613 --> 00:05:25,933 Speaker 4: can do an internet Yeah, you kind of. 93 00:05:26,013 --> 00:05:29,533 Speaker 3: I use cloud storage every day. Does that mean I'm trained? 94 00:05:30,733 --> 00:05:34,173 Speaker 4: Well, you'd probably have to go onto a ws's website 95 00:05:34,253 --> 00:05:37,533 Speaker 4: and do a tutorial or something like that. So there'll 96 00:05:37,533 --> 00:05:40,133 Speaker 4: be a lot of people in that boat. But the 97 00:05:40,173 --> 00:05:42,013 Speaker 4: reality is, you know, we've got a bit of a 98 00:05:42,053 --> 00:05:44,893 Speaker 4: crisis in our education system at the moment. But the 99 00:05:44,933 --> 00:05:48,733 Speaker 4: reality is if you can go now out of high school, 100 00:05:48,933 --> 00:05:51,773 Speaker 4: you can do an eight week or twelve week course 101 00:05:52,293 --> 00:05:56,653 Speaker 4: and become proficient in the entry level stuff to work 102 00:05:57,053 --> 00:06:00,333 Speaker 4: on cloud platforms, and that potentially is a pathway to 103 00:06:00,333 --> 00:06:03,493 Speaker 4: a seventy or eighty thousand dollars job. So I think 104 00:06:03,493 --> 00:06:05,413 Speaker 4: there is something in this that we get hung up 105 00:06:05,453 --> 00:06:09,493 Speaker 4: about tertiary qualifications and that when actually know these sorts 106 00:06:09,533 --> 00:06:13,973 Speaker 4: of companies salesforces, a lot of them offering free qualifications 107 00:06:14,013 --> 00:06:16,413 Speaker 4: and certifications that you can get. The flip side of 108 00:06:16,413 --> 00:06:19,253 Speaker 4: that is that they want you tied to their platform. 109 00:06:19,333 --> 00:06:23,173 Speaker 4: They want you an expert in Salesforce or SAP or Amazon. 110 00:06:23,453 --> 00:06:27,933 Speaker 4: They don't care about giving you general artificial intelligence skills. 111 00:06:28,253 --> 00:06:31,213 Speaker 4: It's really about building a workforce for their platform. Yeah. 112 00:06:31,213 --> 00:06:33,813 Speaker 2: Interesting, there's been a lot of concern about this deal 113 00:06:33,853 --> 00:06:36,413 Speaker 2: when the energy consumption and whether we've got the energy 114 00:06:36,413 --> 00:06:39,293 Speaker 2: infrastructure to supply the likes of these data centers. Is 115 00:06:39,293 --> 00:06:42,213 Speaker 2: that a fair criticism and could that be why the 116 00:06:42,253 --> 00:06:44,813 Speaker 2: infrastructure of these data centers has been holded. 117 00:06:46,413 --> 00:06:49,933 Speaker 4: Yeah. I don't think it's altered them or led to 118 00:06:49,973 --> 00:06:52,613 Speaker 4: any delays, but it is a valid issue because the 119 00:06:52,653 --> 00:06:55,653 Speaker 4: government is basically when they look at the future energy 120 00:06:55,773 --> 00:06:59,493 Speaker 4: makeup of New Zealand and usage, they're saying by twenty thirty, 121 00:06:59,533 --> 00:07:03,213 Speaker 4: potentially up to seven percent off energy use could be 122 00:07:03,653 --> 00:07:07,493 Speaker 4: for data centers. That shows just how this digital economy 123 00:07:07,573 --> 00:07:10,493 Speaker 4: is growing for quickly. There is a lot of renewable 124 00:07:10,653 --> 00:07:14,133 Speaker 4: energy coming online and it's sort of notable that AWS 125 00:07:14,173 --> 00:07:19,573 Speaker 4: has done a specific deal with Mercury to use wind 126 00:07:19,893 --> 00:07:24,573 Speaker 4: energy from the Turretia South wind farm. So they've done 127 00:07:24,893 --> 00:07:28,573 Speaker 4: a long term deal to get guaranteed access to energy. 128 00:07:29,213 --> 00:07:31,973 Speaker 4: Contact Energy did a similar deal with Microsoft, a ten 129 00:07:32,053 --> 00:07:35,213 Speaker 4: year deal. So these big tech companies have come in here, 130 00:07:35,253 --> 00:07:41,053 Speaker 4: they've locked up deals with these energy companies, and the 131 00:07:41,173 --> 00:07:44,253 Speaker 4: energy companies are basically saying this is essential to us 132 00:07:44,333 --> 00:07:48,253 Speaker 4: commissioning new wind farms solar farms. If we have big 133 00:07:48,333 --> 00:07:51,893 Speaker 4: anchor tenants in the form of these tech companies, that 134 00:07:52,093 --> 00:07:55,333 Speaker 4: gives us the confidence to raise investment and build more 135 00:07:55,373 --> 00:07:59,413 Speaker 4: of these. The question is we're already very strained in 136 00:07:59,493 --> 00:08:03,773 Speaker 4: our energy capacity. If we have a winter where the 137 00:08:03,893 --> 00:08:06,173 Speaker 4: lake levels are very low, where there is much wind 138 00:08:06,373 --> 00:08:09,093 Speaker 4: and people are using energy in the evening during the winter, 139 00:08:10,133 --> 00:08:13,573 Speaker 4: is this going to put pressure on our overall grid 140 00:08:13,653 --> 00:08:16,133 Speaker 4: And potentially that is the case. Unless we can get 141 00:08:16,213 --> 00:08:19,573 Speaker 4: enough capacity that's going to be coming online in the 142 00:08:19,573 --> 00:08:22,773 Speaker 4: next five to ten years to absorb this new demand 143 00:08:22,773 --> 00:08:23,533 Speaker 4: from data. 144 00:08:23,253 --> 00:08:25,933 Speaker 3: Centers, well, surely all hands should be the will to 145 00:08:25,973 --> 00:08:28,493 Speaker 3: do that. We should be a nation within all our 146 00:08:28,493 --> 00:08:32,653 Speaker 3: assets of the hydropower, etc. That we can build further 147 00:08:32,813 --> 00:08:35,492 Speaker 3: support to the grid such that huge companies can come 148 00:08:35,533 --> 00:08:37,973 Speaker 3: in here and invest. If we don't do that, then 149 00:08:37,973 --> 00:08:41,173 Speaker 3: we're kind of stuck in a cycle of nothing big 150 00:08:41,213 --> 00:08:45,252 Speaker 3: ever happening. When it comes to the data, Peter, is 151 00:08:45,293 --> 00:08:48,333 Speaker 3: it just New Zealand data that they're looking at storing 152 00:08:48,372 --> 00:08:50,893 Speaker 3: here or is it data from all around the world. 153 00:08:50,933 --> 00:08:54,172 Speaker 3: And if that's the case, what is our capacity for 154 00:08:54,492 --> 00:08:56,012 Speaker 3: bringing data in and out of the country. 155 00:08:57,173 --> 00:09:03,012 Speaker 4: Yeah, Look, I think the actual physical types that transfer 156 00:09:03,213 --> 00:09:05,453 Speaker 4: data in and out of the country is an issue. 157 00:09:05,453 --> 00:09:08,732 Speaker 4: There's lots of capacity there. You've got a new players 158 00:09:08,732 --> 00:09:12,213 Speaker 4: that have plans for even more subse cable, so I 159 00:09:12,213 --> 00:09:15,213 Speaker 4: don't think that's an issue. If you're a New Zealand company. 160 00:09:15,892 --> 00:09:19,813 Speaker 4: One of the big pitches here is not only are 161 00:09:19,852 --> 00:09:22,013 Speaker 4: you closer to a data center, so if you're doing 162 00:09:22,132 --> 00:09:26,052 Speaker 4: a real time service like a chatbot, you know the 163 00:09:26,773 --> 00:09:29,653 Speaker 4: delay or lag in getting an answer back will be 164 00:09:29,693 --> 00:09:32,093 Speaker 4: lower because it's just going to Auckland for it rather 165 00:09:32,093 --> 00:09:36,693 Speaker 4: than Silicon Valley or Australia. So that's a pitch. But 166 00:09:36,813 --> 00:09:39,653 Speaker 4: also you can keep all of your data within New 167 00:09:39,732 --> 00:09:43,372 Speaker 4: Zealand's borders. So a lot of companies and government departments 168 00:09:43,773 --> 00:09:46,372 Speaker 4: are a bit leary about sending a lot of sensitive 169 00:09:46,453 --> 00:09:52,132 Speaker 4: data overseas to other jurisdictions. So aws is saying, we 170 00:09:52,173 --> 00:09:54,493 Speaker 4: can guarantee all the data if you want it to 171 00:09:54,533 --> 00:09:57,933 Speaker 4: stay within New Zealand, we can guarantee that you will 172 00:09:57,973 --> 00:09:58,252 Speaker 4: do that. 173 00:09:58,892 --> 00:10:01,733 Speaker 3: Why do you think there's everyone's lining up to be 174 00:10:01,773 --> 00:10:05,132 Speaker 3: negative about this thing across the media. There's a lot 175 00:10:05,173 --> 00:10:08,172 Speaker 3: of stones being thrown at the thing. Do you think 176 00:10:08,173 --> 00:10:10,132 Speaker 3: it's it's a positive thing and what do you think 177 00:10:10,132 --> 00:10:11,293 Speaker 3: people are being negative about it? 178 00:10:12,012 --> 00:10:16,012 Speaker 4: I think overall, you know, we're in an environment where 179 00:10:16,573 --> 00:10:19,013 Speaker 4: and we've just seen discussion of this with the you know, 180 00:10:19,053 --> 00:10:22,533 Speaker 4: the five million dollar home threshold that Winston Peters has 181 00:10:23,213 --> 00:10:27,492 Speaker 4: has struck a deal around. It's really difficult to get 182 00:10:27,612 --> 00:10:31,093 Speaker 4: companies and high net worth individuals to invest in New 183 00:10:31,213 --> 00:10:36,492 Speaker 4: Zealand where a developed country, but we just don't fit 184 00:10:36,533 --> 00:10:39,133 Speaker 4: their investment profile a lot of them. So anything where 185 00:10:39,132 --> 00:10:42,973 Speaker 4: we're seeing literally billions of dollars come in an investment, 186 00:10:43,132 --> 00:10:45,933 Speaker 4: that is a really good thing. But the worm has 187 00:10:45,973 --> 00:10:48,213 Speaker 4: turned a little bit on on big tech when it 188 00:10:48,213 --> 00:10:51,693 Speaker 4: comes to these data centers because it really is shoring 189 00:10:51,813 --> 00:10:57,693 Speaker 4: up that's sort of cozy duopoly or you know, oligopoly 190 00:10:57,813 --> 00:11:01,413 Speaker 4: that these companies have over our data and increasingly over 191 00:11:01,453 --> 00:11:03,333 Speaker 4: our energy supply as well, and a lot of people 192 00:11:03,413 --> 00:11:06,252 Speaker 4: are like, where is the sovereignty. How do we steer 193 00:11:06,333 --> 00:11:08,653 Speaker 4: our own destiny rather than just give all of our 194 00:11:09,093 --> 00:11:11,933 Speaker 4: valuable data to these three or four big American companies. 195 00:11:12,132 --> 00:11:15,852 Speaker 3: Yeah, yeah, interesting. So just to clarify, this will primarily 196 00:11:15,892 --> 00:11:19,813 Speaker 3: be New Zealand data as opposed to you know, just 197 00:11:21,293 --> 00:11:26,813 Speaker 3: increasing capacity for Amazon's data storage around the world and usage. 198 00:11:27,252 --> 00:11:29,892 Speaker 4: Yeah, well, it's not efficient for them to be sending 199 00:11:30,053 --> 00:11:33,012 Speaker 4: you loads of data from other parts of the world here. 200 00:11:33,053 --> 00:11:36,733 Speaker 4: This will be primarily for New Zealand users, and that's 201 00:11:36,773 --> 00:11:38,853 Speaker 4: going to be the big question. You know, Amazon did 202 00:11:38,892 --> 00:11:41,852 Speaker 4: four hundred and twenty odd million in revenue in New 203 00:11:41,933 --> 00:11:45,292 Speaker 4: Zealand last year. Is there actually, given this sluggish economy, 204 00:11:45,453 --> 00:11:47,693 Speaker 4: enough growth for them to make a decent return on 205 00:11:47,773 --> 00:11:49,532 Speaker 4: this in the next few years. That's going to be 206 00:11:49,573 --> 00:11:49,973 Speaker 4: the question. 207 00:11:50,252 --> 00:11:53,093 Speaker 2: Yeah, Peter, really good to get your analysis and thoughts. 208 00:11:53,132 --> 00:11:54,693 Speaker 2: Thanks so much for joining us. 209 00:11:55,132 --> 00:11:55,852 Speaker 4: Thanks very much. 210 00:11:55,933 --> 00:12:00,213 Speaker 2: That is Peter Griffin, tech commentator and some interesting thoughts there. 211 00:12:00,213 --> 00:12:01,093 Speaker 2: So ken you hear from you. 212 00:12:01,533 --> 00:12:04,213 Speaker 1: For more from News Talks at b Listen live on 213 00:12:04,293 --> 00:12:07,252 Speaker 1: air or online, and keep our shows with you wherever 214 00:12:07,293 --> 00:12:09,852 Speaker 1: you go with our pot casts on iHeartRadio