1 00:00:00,000 --> 00:00:05,340 The AI boom means that data centers are being built at an unprecedented pace around the world 2 00:00:05,340 --> 00:00:09,700 and Isaac, one country in Asia where the build-out is really obvious is Malaysia. 3 00:00:10,140 --> 00:00:14,040 That's right, Mariko. Malaysia has emerged as an AI hub in Southeast Asia. 4 00:00:14,520 --> 00:00:19,760 The country produces a lot of components that go into chips that help power AI technology. 5 00:00:20,220 --> 00:00:23,580 The country has also been aggressively building data centers. 6 00:00:24,080 --> 00:00:27,540 Now, the economy in Malaysia is expanding thanks to this boom, 7 00:00:27,540 --> 00:00:29,860 but there's some pushback too. 8 00:00:30,000 --> 00:00:35,000 Yeah, and the big question is whether or not those concerns about data center are justified. 9 00:00:35,320 --> 00:00:41,680 And also, how is the pushback against them affecting Southeast Asia's effort to get a slice of the AI boom? 10 00:00:45,540 --> 00:00:51,160 Every application you use, whether it's with AI or without AI, has to come from somewhere. 11 00:00:51,500 --> 00:00:54,740 It's not just a little computation that's done in the cloud, right? 12 00:00:55,000 --> 00:00:58,980 The cloud is actually hosted in a physical location called a data center. 13 00:00:58,980 --> 00:01:12,580 One of the things that in these data centers they have is that they will have a lot of network cables, a lot of what they call server racks that is responsible for the computation, and then also the cooling, the energy requirements on there. 14 00:01:12,680 --> 00:01:15,700 And usually the facilities are managed by very, very few people. 15 00:01:15,700 --> 00:01:28,840 So the reason why that has been going was because the compute usage has gone up significantly, probably powered a lot by ChatGPT, Cloud, or every other AI application out there. 16 00:01:28,980 --> 00:01:35,440 And that is why we need to, because we need data centers and we need large sources of land, 17 00:01:35,660 --> 00:01:39,660 you're going to see large sources of construction, there is changes in the environment, 18 00:01:40,440 --> 00:01:44,560 changes in maybe many normal people's daily expenses. 19 00:01:44,800 --> 00:01:46,280 So that's why it's now become a big subject. 20 00:01:46,560 --> 00:01:46,960 Yeah, indeed. 21 00:01:47,360 --> 00:01:52,060 And yeah, because I was looking at, I think in the UK, there were some protests against 22 00:01:52,060 --> 00:01:55,840 those data centers being built in your backyard, also in the United States. 23 00:01:55,840 --> 00:02:02,240 But Isaac, here in Asia, Malaysia kind of stands out in terms of the number of data centers being built. 24 00:02:02,700 --> 00:02:05,960 How did Malaysia become such a hub in terms of this AI boom? 25 00:02:06,040 --> 00:02:11,480 I think the big contributor is back in 2019 when Singapore implemented this temporary ban. 26 00:02:11,620 --> 00:02:15,220 So developers had to look elsewhere to put their data centers. 27 00:02:15,460 --> 00:02:18,000 And so they looked into neighboring Malaysia, Johor. 28 00:02:18,280 --> 00:02:23,180 Its proximity to Singapore makes it a perfect place to build data centers, 29 00:02:23,180 --> 00:02:28,120 which you can see now, 85% of Malaysia's data centres are located in Johor. 30 00:02:28,420 --> 00:02:31,640 Furthermore, the government put out incentives 31 00:02:31,640 --> 00:02:35,460 that sort of draw in foreign investors to invest in the country. 32 00:02:35,840 --> 00:02:41,420 So Bernard, Isaac just talked about Singapore's restrictions on data centres back in 2019. 33 00:02:41,540 --> 00:02:43,060 Can you just explain why that was? 34 00:02:43,120 --> 00:02:45,600 It's actually not so much a ban, it's kind of a moratorium 35 00:02:45,600 --> 00:02:49,060 because I think the Singapore government is very prudent 36 00:02:49,060 --> 00:02:52,420 about how they're thinking about the electricity 37 00:02:52,420 --> 00:02:55,040 outages that have been going on. 38 00:02:55,140 --> 00:02:57,360 They're mostly caused because of a spike in demand 39 00:02:57,360 --> 00:02:58,040 in computing. 40 00:02:58,760 --> 00:03:01,840 And I think post 2024, 41 00:03:02,420 --> 00:03:05,500 when there was also the US banning China 42 00:03:05,500 --> 00:03:07,060 for having the NVIDIA chips 43 00:03:07,060 --> 00:03:09,060 and a lot of the Chinese founders 44 00:03:09,060 --> 00:03:11,600 from the Chinese tech companies, 45 00:03:11,700 --> 00:03:13,280 they require another place 46 00:03:13,280 --> 00:03:14,220 where they could actually have 47 00:03:14,220 --> 00:03:16,120 a source of good compute. 48 00:03:16,540 --> 00:03:18,700 So that also drives in a lot of foreign director 49 00:03:18,700 --> 00:03:24,760 investments into Malaysia as a result. So it's a combination of factors. We also talk about, say, 50 00:03:24,900 --> 00:03:32,020 things like legislation changes. Somewhere between 2019 to 2023, a lot of the Southeast Asia 51 00:03:32,020 --> 00:03:36,600 countries started to have what we call data sovereignty laws. So data sovereignty laws say, 52 00:03:37,020 --> 00:03:41,980 for example, Malaysia said we need to have something like the data must be hosted in Malaysia. 53 00:03:42,740 --> 00:03:46,860 Indonesia certainly say the same. So it also prompts the hyperscalers to say, okay, 54 00:03:46,860 --> 00:03:52,340 we probably need in order to allow our businesses to operate in these countries, 55 00:03:52,520 --> 00:03:54,860 we have to set up the data centers. 56 00:03:55,040 --> 00:03:59,460 And that build up, I think, started as of when that period started. 57 00:03:59,660 --> 00:04:03,320 And I have to say, they don't look that pretty to have in your back garden. 58 00:04:03,620 --> 00:04:05,660 But Isaac, can you explain? 59 00:04:06,300 --> 00:04:09,400 I mean, it's not just geography of where Malaysia was, right? 60 00:04:09,460 --> 00:04:12,240 That all the investments started pouring into the country 61 00:04:12,240 --> 00:04:16,380 because it has been producing a lot of components, tech components. 62 00:04:16,860 --> 00:04:21,800 So how have all these data centers being built? How has it helped the economy? 63 00:04:21,980 --> 00:04:24,940 And can you just explain the background of Malaysia's economy? 64 00:04:25,320 --> 00:04:28,480 Yeah, like billions of dollars are being poured into Malaysia's economy. 65 00:04:28,880 --> 00:04:37,360 And according to the International Monetary Fund, this year alone, it's expected for Malaysia's GDP to grow by at least 4.7%. 66 00:04:37,360 --> 00:04:42,320 Now, the government obviously has a dream of being at the frontier of this AI revolution. 67 00:04:42,320 --> 00:04:45,980 So it's investing a lot. It's offering eligible companies incentives. 68 00:04:45,980 --> 00:04:50,580 For example, some of them could apply for income tax exemptions. 69 00:04:51,000 --> 00:04:54,420 Others could even apply, could get like investment allowances. 70 00:04:54,780 --> 00:04:59,660 So all these incentives are the government's steps to make that dream a reality. 71 00:05:00,380 --> 00:05:05,340 But of course, there has been quite a lot of protests and you actually spoke to somebody. 72 00:05:05,460 --> 00:05:09,360 So can you just tell us what she told you, what her concerns are? 73 00:05:09,780 --> 00:05:11,340 Yeah, I spoke to Itha Malayana. 74 00:05:11,480 --> 00:05:13,280 She's a 26-year-old English teacher. 75 00:05:13,280 --> 00:05:18,000 So in July, she saw the news that a data center was being proposed in her community. 76 00:05:18,560 --> 00:05:21,700 And she basically just said, no, not in my backyard. 77 00:05:22,120 --> 00:05:28,080 And that concern is because she saw what's happening in the US, the huge backlash against 78 00:05:28,080 --> 00:05:31,280 data centers, and she didn't want that to be in her community. 79 00:05:31,400 --> 00:05:34,080 So she goes online and she creates this petition. 80 00:05:34,460 --> 00:05:39,140 And what really stood out for me is that in that petition, she started with a really famous 81 00:05:39,140 --> 00:05:45,080 quote about how when the last tree falls, people will realize that they can't eat money. 82 00:05:45,080 --> 00:05:46,180 And I asked her about that. 83 00:05:46,180 --> 00:05:51,600 She said that quote basically encapsulates everything that she's standing against the 84 00:05:51,600 --> 00:05:53,020 data centers. 85 00:05:53,020 --> 00:05:54,660 And this is what she has to say. 86 00:05:54,660 --> 00:06:01,060 One of the justifications of data centers is that local industry workers would get trained 87 00:06:01,060 --> 00:06:04,480 and hopefully get to develop this industry in Malaysia. 88 00:06:04,480 --> 00:06:09,560 But I think I've read very briefly on this, but it seems like that's not going to be the 89 00:06:09,560 --> 00:06:10,220 case at all. 90 00:06:10,320 --> 00:06:14,140 Like they're just going to set up these servers and then have a few people managing it. 91 00:06:14,200 --> 00:06:16,680 But it's not really going to be like a skill transfer thing. 92 00:06:17,140 --> 00:06:22,920 I could justify using AI in like medicine or like, I don't know, astrophysics. 93 00:06:22,920 --> 00:06:29,040 But why do we need it to decide what we want for lunch or solve a relationship problem? 94 00:06:29,400 --> 00:06:31,660 Bernard, what do you think about what Aoife said? 95 00:06:31,660 --> 00:06:37,660 I think there are a few parts of what it comes to when a data center comes online. 96 00:06:38,240 --> 00:06:43,100 So before the data center is actually built, that actually brings in what we call construction jobs. 97 00:06:43,540 --> 00:06:46,640 Now, once the data center is complete and completed, 98 00:06:47,080 --> 00:06:51,520 then the management of the data center is taken over by, say, the hyperscaler. 99 00:06:51,800 --> 00:06:55,160 They typically, it's actually very few people, as she rightfully pointed. 100 00:06:55,160 --> 00:07:09,180 So what happens here is that then you have a few workers that is actually able to, if there's any anomalous things happening within the data center, they will be the person to resolve the issues on that. 101 00:07:09,480 --> 00:07:14,580 But then when you look out to the bigger economy, right, then you think about what the data center is actually driving. 102 00:07:14,940 --> 00:07:20,000 The data center, the real focus is actually to drive more what we call the digital economy. 103 00:07:20,560 --> 00:07:23,420 So the digital economy could be in the form of, say, 104 00:07:23,920 --> 00:07:26,040 maybe knowledge-based work. 105 00:07:26,160 --> 00:07:29,900 It could be new kinds of jobs that may be created out of this AI. 106 00:07:30,000 --> 00:07:32,320 I know we are now in what we call a transition stage. 107 00:07:32,900 --> 00:07:39,300 And so the big fear is the job displacement is larger than the job creation. 108 00:07:40,000 --> 00:07:40,520 Yeah. 109 00:07:40,780 --> 00:07:44,460 The only answer to what she's saying is the creation of new jobs. 110 00:07:44,620 --> 00:07:46,540 Is there going to be more than or less? 111 00:07:47,120 --> 00:07:47,960 That question, I don't know. 112 00:07:47,960 --> 00:07:52,800 I would prefer to be much more humble to say I don't know. 113 00:07:53,020 --> 00:07:58,640 But I think we've seen different iterations of technological change. 114 00:08:00,000 --> 00:08:02,260 And so far, every time it happened, 115 00:08:02,420 --> 00:08:04,240 we seem to end up creating more jobs. 116 00:08:04,800 --> 00:08:07,920 If you're just looking at this current time period, 117 00:08:08,020 --> 00:08:08,960 when's the data center built? 118 00:08:09,540 --> 00:08:10,820 She's effectively right on that. 119 00:08:10,820 --> 00:08:13,600 But how the data center creates value 120 00:08:13,600 --> 00:08:15,540 for the digital economy within that country, 121 00:08:16,120 --> 00:08:17,700 that is not so clear to me. 122 00:08:17,800 --> 00:08:19,520 But would that be within that country? 123 00:08:19,760 --> 00:08:22,880 Because couldn't you argue that those new jobs created, 124 00:08:23,000 --> 00:08:26,180 thanks to the AI boom, might be outside of Malaysia? 125 00:08:26,280 --> 00:08:29,200 So that was the whole reason why every country 126 00:08:29,200 --> 00:08:30,880 now has their own Data Sovereignty Act. 127 00:08:31,120 --> 00:08:34,100 Yes, you can host other data from other parts of the world, 128 00:08:34,100 --> 00:08:37,700 but you still have a sustainable holding. 129 00:08:38,120 --> 00:08:40,280 And every business out there who needs, 130 00:08:40,380 --> 00:08:41,700 whether it's a neocloud or hyperscaler, 131 00:08:41,700 --> 00:08:46,500 they have to invest in Malaysia in order to get that part of the economy. 132 00:08:46,720 --> 00:08:48,260 And they have to drive the local economy 133 00:08:48,260 --> 00:08:52,300 because what they don't really want is having everyone, 134 00:08:52,560 --> 00:08:55,140 the local communities holding pitchforks and say, 135 00:08:55,240 --> 00:08:57,180 hey, you know, we want to burn down the data center. 136 00:08:57,280 --> 00:09:00,140 That is also a cause of concern for them. 137 00:09:00,600 --> 00:09:03,500 And I think that stability also becomes part of the conversation. 138 00:09:04,080 --> 00:09:08,540 How has the Malaysian government been speaking about this ambition, 139 00:09:08,540 --> 00:09:14,860 their ambition to be the AI hub, but also concerns from those local residents over jobs, 140 00:09:14,860 --> 00:09:20,460 but also about the environment. I previously mentioned that Johor currently houses about 80 141 00:09:20,460 --> 00:09:26,060 to 85% of the data centers in Malaysia. So in the state of Johor, the government has banned two tiers 142 00:09:26,060 --> 00:09:31,900 of data centers. First is a tier one and tier two. Those are data centers which consume about 50 143 00:09:31,900 --> 00:09:39,820 million litres of water a day. And that amount is basically about, it could satisfy the people's 144 00:09:39,820 --> 00:09:47,000 daily drinking needs above 25 million people. So the government has stopped those data centres 145 00:09:47,000 --> 00:09:52,500 from being built. It has also introduced penalties where if the data centre operator said that 146 00:09:52,500 --> 00:09:58,620 this is the amount of electricity it's going to use per year, if it doesn't need 85% of it, 147 00:09:58,620 --> 00:10:00,540 they're going to find the operator. 148 00:10:00,800 --> 00:10:03,220 There's actually a metric to look at data centers. 149 00:10:03,340 --> 00:10:05,220 It's called power usage effectiveness, 150 00:10:05,580 --> 00:10:06,840 what they call the PUE metric. 151 00:10:07,480 --> 00:10:09,720 So if you think about, say, 152 00:10:10,020 --> 00:10:12,240 currently Singapore is mandating 153 00:10:12,240 --> 00:10:15,160 that all data centers must be below 154 00:10:15,160 --> 00:10:17,400 what is called 1.25 PUE. 155 00:10:17,920 --> 00:10:18,000 Okay. 156 00:10:18,580 --> 00:10:20,220 Malaysia, I think it's 1.4. 157 00:10:20,560 --> 00:10:22,460 So there are two aspects of the problem. 158 00:10:22,560 --> 00:10:23,820 So one of them is the water problem. 159 00:10:24,260 --> 00:10:26,120 The water problem is real 160 00:10:26,120 --> 00:10:31,360 if they do not try to recycle the water for the data center's cooling. 161 00:10:31,820 --> 00:10:36,200 So there are actually already known solutions to that. 162 00:10:36,400 --> 00:10:38,360 So there is something called StatePoint Liquid Cooling. 163 00:10:38,800 --> 00:10:40,880 A couple of companies actually started in the US. 164 00:10:41,160 --> 00:10:42,160 But how do they do it? 165 00:10:42,160 --> 00:10:44,120 So basically, you reflux the water. 166 00:10:44,600 --> 00:10:47,960 There is an entire recycling technology on that. 167 00:10:48,420 --> 00:10:51,780 The more concerning part of this is actually energy. 168 00:10:52,900 --> 00:10:55,460 Because of the usage of AI is exponentially growing, 169 00:10:55,460 --> 00:10:58,180 the energy usage is also growing exponentially 170 00:10:58,180 --> 00:11:00,860 and we are actually filling it. 171 00:11:01,200 --> 00:11:05,640 So a lot of it is due to poor demand forecasting of power. 172 00:11:05,840 --> 00:11:07,420 If you think about, look at the region as a whole, 173 00:11:07,740 --> 00:11:09,940 countries, I mean, Singapore is already doing that. 174 00:11:10,160 --> 00:11:11,580 The next data centre you build, 175 00:11:11,700 --> 00:11:13,640 you have to have a certain energy, 176 00:11:14,820 --> 00:11:16,060 that means the energy you use, 177 00:11:16,240 --> 00:11:18,080 some parts of it must come from renewables. 178 00:11:18,560 --> 00:11:20,580 Then as you go to say Malaysia, for example, 179 00:11:20,720 --> 00:11:21,860 then the question is what kind of, 180 00:11:22,160 --> 00:11:23,520 they definitely have energy, 181 00:11:24,000 --> 00:11:25,100 they have petroleum, oil, 182 00:11:25,460 --> 00:11:29,860 so they can hydrocarbon so you can power some part of it then the question is do they want to 183 00:11:29,860 --> 00:11:36,820 rely on other mixes because the there is no way now that uh just by using a carbon they are able 184 00:11:36,820 --> 00:11:42,420 to run that they also you also bring about the increase of climate change as well and malaysia 185 00:11:42,420 --> 00:11:47,620 has the ambition to be that ai hub in southeast asia and it really wants to achieve that so i 186 00:11:47,620 --> 00:11:52,420 think the government needs to balance between addressing the people's concerns while also 187 00:11:52,420 --> 00:11:58,180 taking realistic steps to make those dreams a reality of being an AI hub. So the Malaysian 188 00:11:58,180 --> 00:12:04,660 government, the Prime Minister announced that the young people will be able to access a three-month 189 00:12:04,660 --> 00:12:10,500 free subscription to some of the AI models and that is basically to allow them to explore the 190 00:12:10,500 --> 00:12:17,460 AI models, play with it and even to perhaps boost their creativity and improve their digital skills. 191 00:12:17,460 --> 00:12:22,520 So the government is taking an active step into making the country embrace AI. 192 00:12:22,740 --> 00:12:26,980 But I guess we'll see how the government addresses the people's concerns. 193 00:12:27,760 --> 00:12:33,180 Will they keep building data centers for now with, I guess, stricter regulations, Isaac? 194 00:12:33,440 --> 00:12:37,200 I think so, honestly, because the people's concerns about data centers are valid. 195 00:12:37,200 --> 00:12:43,600 And for example, you heard from Ether herself that she said that she doesn't want Malaysia to become like the US, 196 00:12:43,600 --> 00:12:47,160 where the people actively push back against data centers 197 00:12:47,160 --> 00:12:52,460 because she's questioning why is data centers, AI, relevant or even important? 198 00:12:52,460 --> 00:12:56,420 Because we've heard from the AI company CEOs themselves 199 00:12:56,420 --> 00:12:59,100 that they say that this AI revolution is inevitable. 200 00:12:59,620 --> 00:13:00,980 But she's questioning that. 201 00:13:01,320 --> 00:13:05,580 So I think we'll hear more and more people having similar concerns as Ether. 202 00:13:06,000 --> 00:13:09,700 One interesting example of a country that's in this AI revolution 203 00:13:09,700 --> 00:13:12,560 that's not very much we would think about is China. 204 00:13:12,560 --> 00:13:15,820 so where China is pretty effective 205 00:13:15,820 --> 00:13:17,520 is the energy infrastructure 206 00:13:17,520 --> 00:13:18,480 so they have built up 207 00:13:18,480 --> 00:13:20,060 most of the energy infrastructure 208 00:13:20,060 --> 00:13:21,300 they have built up 209 00:13:21,300 --> 00:13:22,080 most of the 210 00:13:22,080 --> 00:13:23,460 renewable energy mix 211 00:13:23,460 --> 00:13:24,620 what they don't have 212 00:13:24,620 --> 00:13:25,340 is now the chips 213 00:13:25,340 --> 00:13:27,160 and the AI models are lagging 214 00:13:27,160 --> 00:13:28,820 basically 3 to 6 months 215 00:13:28,820 --> 00:13:30,060 depending on who you talk to 216 00:13:30,060 --> 00:13:31,120 on that 217 00:13:31,120 --> 00:13:32,820 but if you look at say 218 00:13:32,820 --> 00:13:34,400 there was a survey done 219 00:13:34,400 --> 00:13:36,420 80% of Chinese 220 00:13:36,420 --> 00:13:38,320 are hopeful about the AI future 221 00:13:38,320 --> 00:13:40,600 and then if you do the same survey 222 00:13:40,600 --> 00:13:42,080 in the US it's only 31% 223 00:13:42,080 --> 00:13:44,180 So the question is, what's the big difference, right? 224 00:13:44,440 --> 00:13:48,300 So I think a lot of it is in whether we're optimistic 225 00:13:48,300 --> 00:13:52,820 that the AI will bring about sort of a growth of a new economy. 226 00:13:53,660 --> 00:13:59,520 And there are some things that I think people take it to the logical extreme. 227 00:13:59,840 --> 00:14:03,860 If AI was to do everything of this, then what should humans do, right? 228 00:14:04,180 --> 00:14:07,920 I think the way I would probably want to moving ahead, 229 00:14:07,920 --> 00:14:12,260 I think that there is actually human creativity is actually infinite. 230 00:14:12,580 --> 00:14:14,620 The demand for human creativity is actually infinite. 231 00:14:15,440 --> 00:14:20,760 And I think for that, because of that, even if AI can automate most of the things, 232 00:14:21,100 --> 00:14:22,360 there are still things for us to do. 233 00:14:22,640 --> 00:14:23,960 Yeah, really insightful conversation. 234 00:14:24,120 --> 00:14:26,240 Thank you so much for joining me, Bernad and Isaac.