1 00:00:05,120 --> 00:00:06,320 Speaker 1: This is what the fluck. 2 00:00:06,440 --> 00:00:08,640 Speaker 2: I'm Brett and I'm Justin and it's Wednesday, the eighth 3 00:00:08,680 --> 00:00:09,080 Speaker 2: of April. 4 00:00:09,200 --> 00:00:12,639 Speaker 1: Juseyboy. COVID might be old news, but Australian CBDs clearly 5 00:00:12,640 --> 00:00:14,520 Speaker 1: did not get the memo. Get this. The vacancy rate 6 00:00:14,560 --> 00:00:17,400 Speaker 1: in the CBD markets across Australia is fifteen point nine percent, 7 00:00:17,480 --> 00:00:20,079 Speaker 1: which is the highest level since nineteen ninety five according 8 00:00:20,120 --> 00:00:22,720 Speaker 1: to the Property Council of Australia and the gold medal 9 00:00:22,760 --> 00:00:26,200 Speaker 1: for the most vacant CBD that would be Humble Old 10 00:00:26,360 --> 00:00:28,880 Speaker 1: Melbourne at nineteen percent. It's the ghost town around nor 11 00:00:28,920 --> 00:00:31,240 Speaker 1: one c Yeah, fox an. A new month means a 12 00:00:31,280 --> 00:00:33,000 Speaker 1: new chance to check your credit score and see you 13 00:00:33,040 --> 00:00:35,440 Speaker 1: winner it went down or what's changed. So if you 14 00:00:35,440 --> 00:00:36,839 Speaker 1: want to say where the banks can see about you 15 00:00:36,920 --> 00:00:39,160 Speaker 1: before you apply for a loan, make sure to download 16 00:00:39,200 --> 00:00:41,199 Speaker 1: the Flux app and check out your credit score through 17 00:00:41,560 --> 00:00:44,000 Speaker 1: riveting Stories today, jussy boy, let's do it for our 18 00:00:44,080 --> 00:00:48,000 Speaker 1: first Australian data center operator, next DC is issuing a 19 00:00:48,120 --> 00:00:51,159 Speaker 1: one hundred year, one billion dollar bond to fund its 20 00:00:51,280 --> 00:00:52,720 Speaker 1: data center expansion company. 21 00:00:52,760 --> 00:00:55,600 Speaker 2: It's been around for fifteen years, lockxing debt for a century. 22 00:00:56,080 --> 00:00:58,960 Speaker 1: Wowee, tell me more so. Next DC is Australia's largest 23 00:00:59,000 --> 00:01:00,480 Speaker 1: listed data center operator. 24 00:01:00,560 --> 00:01:02,680 Speaker 2: Think of it as the landlord of the Internet. 25 00:01:02,760 --> 00:01:05,800 Speaker 1: It owns seventeen data centers across Australia and Asia, with 26 00:01:06,040 --> 00:01:08,640 Speaker 1: another eleven in the pipeline. And be man data centers 27 00:01:08,680 --> 00:01:11,400 Speaker 1: are having a moment right now. Yeah, forget beachfront property. 28 00:01:11,400 --> 00:01:14,320 Speaker 1: The hottest real estate right now is windowless warehouses full 29 00:01:14,360 --> 00:01:15,759 Speaker 1: of servers and air conditioners. 30 00:01:15,840 --> 00:01:18,640 Speaker 2: Yeah, we've seen demand boom because of AI platforms like Chutch, 31 00:01:18,640 --> 00:01:22,120 Speaker 2: Avert and big tech companies needing more processing power. But 32 00:01:22,160 --> 00:01:25,160 Speaker 2: to expand quickly, NEXTDC needs more more art to build 33 00:01:25,160 --> 00:01:25,959 Speaker 2: those data centers. 34 00:01:25,959 --> 00:01:28,680 Speaker 1: So now, NEXTDC has just locked in a one billion 35 00:01:28,760 --> 00:01:31,200 Speaker 1: dollar debt deal, but not your average one. We're talking 36 00:01:31,240 --> 00:01:34,520 Speaker 1: one hundred year bond back by a Canadian pension giant 37 00:01:34,520 --> 00:01:36,399 Speaker 1: and this new deal gives them the capital they need 38 00:01:36,440 --> 00:01:39,080 Speaker 1: to push ahead with a massive two point seven billion 39 00:01:39,120 --> 00:01:40,360 Speaker 1: dollar data center build out. 40 00:01:40,440 --> 00:01:43,000 Speaker 2: Next minute, next DC share price spikes twelve percent on 41 00:01:43,040 --> 00:01:45,920 Speaker 2: the news because investors like this long term loan. 42 00:01:45,720 --> 00:01:46,280 Speaker 1: Do they really? 43 00:01:46,280 --> 00:01:46,360 Speaker 3: So? 44 00:01:46,440 --> 00:01:47,480 Speaker 1: What is the key learning here? 45 00:01:47,520 --> 00:01:49,800 Speaker 2: A bond is essentially a loan from investors to a 46 00:01:49,840 --> 00:01:52,080 Speaker 2: company or government Instead of going to a bank. 47 00:01:52,200 --> 00:01:54,960 Speaker 1: Companies can raise money by issuing bonds to the public 48 00:01:55,160 --> 00:01:56,600 Speaker 1: or large institutions. 49 00:01:56,640 --> 00:01:58,840 Speaker 2: It's how it works. Investors lend money up front. 50 00:01:58,640 --> 00:02:00,919 Speaker 1: And in return, the company greats to pay regular interests, 51 00:02:01,000 --> 00:02:02,600 Speaker 1: call the coupon and repay the full. 52 00:02:02,480 --> 00:02:04,560 Speaker 2: Amount later, like one hundred years later. 53 00:02:04,920 --> 00:02:07,640 Speaker 1: Companies like NEXTDC use bonds because they can raise large 54 00:02:07,680 --> 00:02:09,280 Speaker 1: amounts of capital quickly. 55 00:02:08,960 --> 00:02:11,840 Speaker 2: And they can avoid giving up ownership unlike isseran shares, and. 56 00:02:11,800 --> 00:02:14,000 Speaker 1: They can lock in funding for long term projects, and. 57 00:02:14,000 --> 00:02:16,400 Speaker 2: They man for investors. Bonds are typically seen as more 58 00:02:16,440 --> 00:02:19,720 Speaker 2: stable than shares because they provide predictable income in Jazwei. 59 00:02:19,760 --> 00:02:22,079 Speaker 1: While one hundred year bonds sound a little bit crazy, 60 00:02:22,280 --> 00:02:25,080 Speaker 1: NEXTDC is not the only one who has done it. 61 00:02:25,360 --> 00:02:27,799 Speaker 2: The World Disney Company issued one hundred year bond back 62 00:02:27,800 --> 00:02:28,919 Speaker 2: in nineteen ninety three. 63 00:02:29,040 --> 00:02:32,040 Speaker 1: Also, Coca Cola followed Disney in selling a one hundred 64 00:02:32,200 --> 00:02:34,320 Speaker 1: year bond, which is where, given how fast AI has 65 00:02:34,360 --> 00:02:36,440 Speaker 1: moved in the last two years, who even knows, because 66 00:02:36,440 --> 00:02:39,480 Speaker 1: the world looks like in one hundred years whole century 67 00:02:40,120 --> 00:02:42,239 Speaker 1: Fox Sam, We're excited to be joined by someone who 68 00:02:42,240 --> 00:02:45,200 Speaker 1: spends his days thinking about where global markets are heading 69 00:02:45,400 --> 00:02:48,080 Speaker 1: and where the real opportunities are hiding. Ben Arnold is 70 00:02:48,120 --> 00:02:51,079 Speaker 1: the Investment director for Global Equities at Schroders, where he 71 00:02:51,120 --> 00:02:54,200 Speaker 1: works across global portfolios and sits alongside some of the 72 00:02:54,200 --> 00:02:57,400 Speaker 1: world's top investors. He's also spent the last decade watching 73 00:02:57,400 --> 00:02:59,960 Speaker 1: how the best in the game actually make investment decisions. 74 00:03:00,120 --> 00:03:01,680 Speaker 1: So Ben, welcome to what the plaques. 75 00:03:01,840 --> 00:03:02,880 Speaker 3: Thanks so much for having me. 76 00:03:03,080 --> 00:03:05,320 Speaker 2: Okay, I'm keen to know because we have seen the 77 00:03:05,400 --> 00:03:08,000 Speaker 2: A six and the nazeg wobble big time off the 78 00:03:08,000 --> 00:03:10,320 Speaker 2: back of the Iran conflict. So I want to know 79 00:03:10,480 --> 00:03:12,480 Speaker 2: do you think this is just a short term scire 80 00:03:12,680 --> 00:03:15,200 Speaker 2: or is this something that investors should be taking seriously 81 00:03:15,240 --> 00:03:15,880 Speaker 2: for the long term. 82 00:03:15,919 --> 00:03:18,720 Speaker 3: Well, that's the question on the market's lips at the minute. 83 00:03:18,800 --> 00:03:21,800 Speaker 3: And given the news flow and things are so volatile 84 00:03:21,960 --> 00:03:24,399 Speaker 3: and changing every day that the market is really trying 85 00:03:24,400 --> 00:03:26,400 Speaker 3: to work out how long is this thing going to last? 86 00:03:26,480 --> 00:03:28,600 Speaker 3: Is Trump going to be able to meet his objectives 87 00:03:28,639 --> 00:03:31,400 Speaker 3: in a very short time a matter of days and weeks, 88 00:03:31,480 --> 00:03:33,440 Speaker 3: or is it going to be a much more prolonged 89 00:03:33,639 --> 00:03:36,120 Speaker 3: period of conflict and then all the consequences of that 90 00:03:36,240 --> 00:03:38,400 Speaker 3: to both you know, his popularity in the US, and 91 00:03:38,440 --> 00:03:42,080 Speaker 3: then obviously the investment consequences, which has fundamental investors is 92 00:03:42,120 --> 00:03:44,160 Speaker 3: what is what we think about, you know, I think 93 00:03:44,320 --> 00:03:47,240 Speaker 3: for investors it's very very important to acknowledge where you 94 00:03:47,240 --> 00:03:49,440 Speaker 3: think you have an edge as an active investor. You know, 95 00:03:49,680 --> 00:03:55,200 Speaker 3: we have access to lots and lots of amazing resource tanks, experts, 96 00:03:55,480 --> 00:03:58,800 Speaker 3: political specialists, and so we try to proceed with a 97 00:03:58,880 --> 00:04:01,520 Speaker 3: kind of milestone approach, looking at evidence of what is 98 00:04:01,520 --> 00:04:05,600 Speaker 3: happening on the ground rather than the noise in the media. So, 99 00:04:06,200 --> 00:04:09,760 Speaker 3: for example, if there was evidence of a civil war 100 00:04:09,880 --> 00:04:13,920 Speaker 3: developing internal fractions or some kind of data point to 101 00:04:13,960 --> 00:04:16,440 Speaker 3: say that there may be boots on the ground, for instance, 102 00:04:16,880 --> 00:04:19,280 Speaker 3: that would suggest this conflict is going to last a 103 00:04:19,320 --> 00:04:22,000 Speaker 3: lot longer. So we want to be data driven and 104 00:04:22,120 --> 00:04:25,239 Speaker 3: milestone driven, but it is very difficult. We would always 105 00:04:25,320 --> 00:04:27,839 Speaker 3: kind of encourage people to think about investing and with 106 00:04:27,920 --> 00:04:30,240 Speaker 3: a long term view. Very few people in history have 107 00:04:30,279 --> 00:04:33,440 Speaker 3: been able to trade around these short events, you know, 108 00:04:33,640 --> 00:04:36,640 Speaker 3: successfully time and time again, So we always try and 109 00:04:36,720 --> 00:04:39,359 Speaker 3: kind of focus on the long term and separate that 110 00:04:39,560 --> 00:04:40,720 Speaker 3: from the noise. Every day. 111 00:04:41,200 --> 00:04:43,560 Speaker 2: You obviously speak a lot about long term investing and 112 00:04:43,839 --> 00:04:46,640 Speaker 2: schro this is an active ATF manager. So what does 113 00:04:46,640 --> 00:04:49,200 Speaker 2: that actually mean and how does that contribute towards the 114 00:04:49,200 --> 00:04:50,800 Speaker 2: Schroeder's investment philosophy. 115 00:04:50,920 --> 00:04:53,720 Speaker 3: Yeah, so we have a couple of active ETFs and 116 00:04:53,760 --> 00:04:56,839 Speaker 3: those are kind of products that are easily accessible to 117 00:04:57,040 --> 00:04:59,600 Speaker 3: lots of lots of retail investors. In Australia, we have too, 118 00:05:00,200 --> 00:05:03,520 Speaker 3: is called Global Equity Alpha. It's under the ticker ALP 119 00:05:03,640 --> 00:05:07,960 Speaker 3: ALPH and that's what you call a fundamental stockpicking fund. 120 00:05:08,240 --> 00:05:11,600 Speaker 3: So you have a far manager making decisions. They're supported 121 00:05:11,640 --> 00:05:15,200 Speaker 3: by a bank of analysts around the world in lots 122 00:05:15,200 --> 00:05:19,719 Speaker 3: of different Troders' offices, speaking to management teams, building models 123 00:05:19,720 --> 00:05:23,680 Speaker 3: of each company and proposing stocks to buy for the 124 00:05:23,760 --> 00:05:27,159 Speaker 3: portfolio manager. So he then brings that together. It's forty 125 00:05:27,160 --> 00:05:29,600 Speaker 3: to sixty stocks and the idea is that that can 126 00:05:29,640 --> 00:05:33,080 Speaker 3: outperform by between two and three percent above the market 127 00:05:33,360 --> 00:05:36,080 Speaker 3: over the medium and long term per year. And that's 128 00:05:36,160 --> 00:05:38,719 Speaker 3: very much a stockpicking fund. And then the other flagship 129 00:05:38,720 --> 00:05:42,320 Speaker 3: strategy is the Schroder Global Core Strategy. The ticker for 130 00:05:42,400 --> 00:05:46,440 Speaker 3: that is Core CEORE and that's a very different approach, 131 00:05:46,480 --> 00:05:49,360 Speaker 3: managed by very different team, but still within Troders, and 132 00:05:49,400 --> 00:05:53,920 Speaker 3: that's what we call a quantitative strategy or enhanced index. 133 00:05:54,279 --> 00:05:57,440 Speaker 3: So the portfolio has a lot more names in it, 134 00:05:57,520 --> 00:05:59,800 Speaker 3: So instead of forty to sixty, we're thinking more sort 135 00:05:59,839 --> 00:06:00,960 Speaker 3: of four to five hundred. 136 00:06:01,160 --> 00:06:01,480 Speaker 1: Wow. 137 00:06:01,560 --> 00:06:03,880 Speaker 3: And the starting point is looking at the universe and 138 00:06:03,920 --> 00:06:08,479 Speaker 3: then basically applying some quantitative methods to say, well, which 139 00:06:08,560 --> 00:06:10,760 Speaker 3: stops within that universe do I want to be slightly 140 00:06:10,839 --> 00:06:14,479 Speaker 3: overweight versus the market or slightly underweight, and then using 141 00:06:14,520 --> 00:06:16,480 Speaker 3: our tools to kind of scale that up. And the 142 00:06:16,520 --> 00:06:19,880 Speaker 3: whole idea is that that portfolio will outperform the market 143 00:06:19,960 --> 00:06:24,359 Speaker 3: by roughly a percent a year before fees, and so 144 00:06:24,440 --> 00:06:27,920 Speaker 3: that's very much more of an enhanced index strategy. So 145 00:06:27,960 --> 00:06:30,400 Speaker 3: those are the two strategies that we were the traders. 146 00:06:30,520 --> 00:06:33,760 Speaker 3: Traders are huge supporters of active management. We've always been 147 00:06:33,800 --> 00:06:37,479 Speaker 3: active investors. We believe that markets are inefficient over the 148 00:06:37,520 --> 00:06:40,479 Speaker 3: long term, and we have a very strong pedigree in 149 00:06:40,560 --> 00:06:43,919 Speaker 3: terms of generating consistent out performance for our clients. 150 00:06:44,080 --> 00:06:44,240 Speaker 1: beIN. 151 00:06:44,360 --> 00:06:46,000 Speaker 2: Thanks so much for joining us on what the Flux. 152 00:06:46,040 --> 00:06:48,480 Speaker 2: Always great to hear your insights, so looking forward to 153 00:06:48,520 --> 00:06:49,200 Speaker 2: checking in soon. 154 00:06:49,279 --> 00:06:51,279 Speaker 3: Thanks so much for. 155 00:06:51,160 --> 00:06:54,520 Speaker 2: Our second story. Safety culture has kicked off an acquisition spree, 156 00:06:54,800 --> 00:06:57,960 Speaker 2: buying a Sydney based AI startup called Twine as it 157 00:06:58,040 --> 00:06:59,920 Speaker 2: gets ready with a scary AI. 158 00:06:59,640 --> 00:07:03,120 Speaker 1: Era Safety Culture, making workplace safety checklist sexy since two 159 00:07:03,120 --> 00:07:04,719 Speaker 1: thousand and four, tell me more so. 160 00:07:04,760 --> 00:07:06,680 Speaker 2: Safety Culture started in two thousand and four as a 161 00:07:06,680 --> 00:07:08,680 Speaker 2: workplace safety checklist act, the. 162 00:07:08,720 --> 00:07:10,920 Speaker 1: Kind of thing that makes sure no one forgets to 163 00:07:10,960 --> 00:07:11,880 Speaker 1: wear a hard. 164 00:07:11,680 --> 00:07:14,640 Speaker 2: Hatch, and since then it's grown into a global workplace 165 00:07:14,720 --> 00:07:16,160 Speaker 2: operations platform. 166 00:07:15,800 --> 00:07:17,400 Speaker 1: Which has valued at about two and a half billion 167 00:07:17,440 --> 00:07:19,720 Speaker 1: dollars at its last rays in September twenty twenty four. 168 00:07:19,760 --> 00:07:21,640 Speaker 2: But the mean, here's the thing, Even a company built 169 00:07:21,640 --> 00:07:24,520 Speaker 2: on safety is feeling a little unsafe in this world 170 00:07:24,560 --> 00:07:25,320 Speaker 2: of AI right now. 171 00:07:25,360 --> 00:07:28,440 Speaker 1: So now Safety Culture is aggressively acquiring smaller companies to 172 00:07:28,440 --> 00:07:31,000 Speaker 1: build out its AI capabilities and stay competitive. 173 00:07:31,040 --> 00:07:33,840 Speaker 2: Yep, it's just acquired a Sydney based AI startup called Twine, 174 00:07:33,920 --> 00:07:37,200 Speaker 2: which uses AI to analyze sales and customer service calls. 175 00:07:37,240 --> 00:07:39,720 Speaker 1: And this isn't a one off. They've already got another 176 00:07:39,720 --> 00:07:42,080 Speaker 1: acquisition lined up and are in late stage talks for 177 00:07:42,120 --> 00:07:42,520 Speaker 1: a third. 178 00:07:42,560 --> 00:07:44,960 Speaker 2: And the Safety Culture is clearly going on the offensive 179 00:07:45,000 --> 00:07:48,160 Speaker 2: as AI begins to reshape the global software as a 180 00:07:48,160 --> 00:07:50,840 Speaker 2: service industry and plot twist. Twine and Safety Culture share 181 00:07:50,880 --> 00:07:52,440 Speaker 2: the same key investor, air Tree. 182 00:07:52,440 --> 00:07:54,440 Speaker 1: Who world's colliding? So what's the key learning here? 183 00:07:54,520 --> 00:07:57,040 Speaker 2: In venture capital, it's actually quite common for companies backed 184 00:07:57,040 --> 00:07:59,320 Speaker 2: by the same investors to end up doing deals with 185 00:07:59,360 --> 00:08:02,720 Speaker 2: each other. Example, Canvas has acquired companies like Leonardo AI 186 00:08:02,920 --> 00:08:05,200 Speaker 2: for three hundred and twenty million bucks. It also acquired 187 00:08:05,240 --> 00:08:07,400 Speaker 2: Magic Brief over twenty two million US bucks. 188 00:08:07,440 --> 00:08:10,200 Speaker 1: Both of those companies were backed by Blackbird, like Canva, 189 00:08:10,280 --> 00:08:12,720 Speaker 1: who also is well backed by Blackbird. 190 00:08:12,280 --> 00:08:15,280 Speaker 2: And now Safety Culture, backed by Airtree, has acquired Twine, 191 00:08:15,320 --> 00:08:17,880 Speaker 2: which also had capital raises led by Airtree. 192 00:08:17,920 --> 00:08:19,800 Speaker 1: Even back in the day, we saw Facebook, which was 193 00:08:19,840 --> 00:08:22,680 Speaker 1: backed by Axcel, acquire Instagram for one billion US dollars, 194 00:08:22,720 --> 00:08:24,800 Speaker 1: which was also backed by Xcel. So be man, tell 195 00:08:24,880 --> 00:08:28,000 Speaker 1: me why does this happen? Well, vcs are actively connecting 196 00:08:28,040 --> 00:08:31,080 Speaker 1: their portfolio companies, and when one company needs capability, the 197 00:08:31,080 --> 00:08:33,240 Speaker 1: first place of VC well look is inside their own 198 00:08:33,280 --> 00:08:35,960 Speaker 1: portfolio and jasboy, it makes sense that it increases the 199 00:08:36,040 --> 00:08:38,400 Speaker 1: chances of successful outcome between the two. 200 00:08:38,800 --> 00:08:41,560 Speaker 2: The investors understand both businesses, they trust the founders, and 201 00:08:41,600 --> 00:08:43,040 Speaker 2: they can help facilitate the deal. 202 00:08:43,080 --> 00:08:45,280 Speaker 1: And when capital is tight or competition is heating up. 203 00:08:45,320 --> 00:08:48,480 Speaker 1: Like with AI, these within network acquisitions can happen even 204 00:08:48,520 --> 00:08:51,880 Speaker 1: more frequently. For our third and final story, AI rivals 205 00:08:51,920 --> 00:08:55,720 Speaker 1: Open Ai, Anthropic and alsbed's Google have begun working together 206 00:08:55,920 --> 00:08:59,640 Speaker 1: to clamp down on Chinese competitors stealing their models. Alrighty, 207 00:08:59,679 --> 00:09:02,839 Speaker 1: this is giving tech cold war energy. I am listening, so. 208 00:09:02,800 --> 00:09:05,000 Speaker 2: We know that AI chatbots are kind of taking over 209 00:09:05,040 --> 00:09:05,440 Speaker 2: our lives. 210 00:09:05,520 --> 00:09:08,439 Speaker 1: Yeah, Between chat GBT and Claude and Gemini and co pilot, 211 00:09:08,440 --> 00:09:10,760 Speaker 1: we're talking more than one billion monthly active users. 212 00:09:10,840 --> 00:09:13,400 Speaker 2: And back in twenty twenty three, Open Ai, Anthropy, Google, 213 00:09:13,400 --> 00:09:16,760 Speaker 2: and Microsoft founded a group called the Frontier Model Forum. 214 00:09:17,120 --> 00:09:19,959 Speaker 2: Very Avengers coded yeah, But instead of holding hands and 215 00:09:20,000 --> 00:09:23,240 Speaker 2: singing Kumbaya, these tech companies mostly just competed with each 216 00:09:23,240 --> 00:09:26,920 Speaker 2: other until now Forum is back baby yep. Open Ai, Google, 217 00:09:26,960 --> 00:09:29,600 Speaker 2: and Anthropic have started to work together to crack down 218 00:09:29,640 --> 00:09:33,160 Speaker 2: on competitors, particularly in China, from copying their AI model. 219 00:09:33,200 --> 00:09:35,280 Speaker 2: But beyond the money, these companies have concerns that this 220 00:09:35,400 --> 00:09:38,200 Speaker 2: free writing could become a national security risk in the US, 221 00:09:38,640 --> 00:09:41,000 Speaker 2: So the big tech and AI companies are working together 222 00:09:41,120 --> 00:09:43,920 Speaker 2: and sharing data to try and crack down on these copycats. 223 00:09:43,960 --> 00:09:47,559 Speaker 1: And crack down on their distillation. So what's the key 224 00:09:47,600 --> 00:09:50,679 Speaker 1: learning here? AI distillation is a technique where a smaller 225 00:09:50,800 --> 00:09:53,600 Speaker 1: or newer model called the student, learns from a larger, 226 00:09:53,679 --> 00:09:55,199 Speaker 1: more advanced model called the teacher. 227 00:09:55,280 --> 00:09:57,200 Speaker 2: So instead of building a model from scratch, which can 228 00:09:57,200 --> 00:10:00,280 Speaker 2: cost billions, developers use outputs from an existing model to 229 00:10:00,360 --> 00:10:04,520 Speaker 2: replicate its capabilities. And in some cases, distillation is completely legit. 230 00:10:04,640 --> 00:10:07,200 Speaker 2: Ye like when companies use it internally to create faster 231 00:10:07,320 --> 00:10:09,360 Speaker 2: and cheaper versions of their older models. 232 00:10:09,360 --> 00:10:10,360 Speaker 1: That makes sense that they man. 233 00:10:10,400 --> 00:10:13,480 Speaker 2: It becomes pretty controversial when third parties use these models 234 00:10:13,480 --> 00:10:14,400 Speaker 2: without permission. 235 00:10:14,600 --> 00:10:15,600 Speaker 1: How do they do this? Well? 236 00:10:15,640 --> 00:10:18,839 Speaker 2: They do this by repeatedly queering a powerful a model 237 00:10:18,880 --> 00:10:20,479 Speaker 2: and collecting all of the responses. 238 00:10:20,559 --> 00:10:22,640 Speaker 1: Let me just keep asking questions until I figure out 239 00:10:22,679 --> 00:10:23,480 Speaker 1: how your brain. 240 00:10:23,320 --> 00:10:25,920 Speaker 2: Works and boom, they built something similar at a fraction 241 00:10:26,040 --> 00:10:26,480 Speaker 2: of the cost. 242 00:10:26,520 --> 00:10:29,280 Speaker 1: And just boy. US officials estimate this kind of activity 243 00:10:29,320 --> 00:10:32,280 Speaker 1: is costing AI companies billions in lost profits each year. 244 00:10:32,360 --> 00:10:32,920 Speaker 1: So be man. 245 00:10:32,960 --> 00:10:35,720 Speaker 2: Even the biggest competitors will collaborate when the threat is 246 00:10:35,720 --> 00:10:38,160 Speaker 2: big enough. Flux Sam, your credit score could be between 247 00:10:38,280 --> 00:10:41,079 Speaker 2: zero and twelve hundred. That is a very big range, 248 00:10:41,080 --> 00:10:42,560 Speaker 2: and you kind of want to know where you stand, 249 00:10:42,600 --> 00:10:43,880 Speaker 2: so if you haven't checked your credit score for a 250 00:10:43,880 --> 00:10:45,559 Speaker 2: few months, now is the perfect time to do it 251 00:10:45,600 --> 00:10:47,440 Speaker 2: in the Flux app, but you can see just cause one. 252 00:10:47,400 --> 00:10:49,079 Speaker 1: Up or down and why it's changed. 253 00:10:49,120 --> 00:10:51,000 Speaker 2: Make sure to download the Flux app and check out 254 00:10:51,000 --> 00:10:51,600 Speaker 2: your credit score. 255 00:10:51,640 --> 00:10:54,120 Speaker 1: Thanks for listening, and we'll see you on Friday.