WEBVTT - Why DeepSeek could be an AI game changer

0:00:00.560 --> 0:00:03.720
<v Speaker 1>Already and this is this is the daily This is

0:00:03.720 --> 0:00:15.160
<v Speaker 1>the Daily OS. Oh, now it makes sense. Good morning

0:00:15.160 --> 0:00:17.439
<v Speaker 1>and welcome to the Daily OS. It's Wednesday, the twenty

0:00:17.520 --> 0:00:20.920
<v Speaker 1>ninth of January. I'm Sam, I'm Zara. What if creating

0:00:21.040 --> 0:00:25.200
<v Speaker 1>artificial intelligence programs was actually cheaper than we thought? How

0:00:25.239 --> 0:00:29.480
<v Speaker 1>would AI companies, governments and investors react if something faster,

0:00:29.800 --> 0:00:33.080
<v Speaker 1>more accurate, and cheaper than the current AI giants like

0:00:33.200 --> 0:00:37.120
<v Speaker 1>chatcha Et suddenly burst onto the scene. Well, today we're

0:00:37.120 --> 0:00:39.479
<v Speaker 1>going to talk about Deep Seek, this new company that

0:00:39.520 --> 0:00:42.320
<v Speaker 1>has really shaken up the stock market and the way

0:00:42.320 --> 0:00:46.280
<v Speaker 1>that we're all feeling about artificial intelligence. It's pretty remarkable,

0:00:46.280 --> 0:00:46.880
<v Speaker 1>stories areah.

0:00:51.000 --> 0:00:53.240
<v Speaker 2>I think at the beginning of this week, the number

0:00:53.240 --> 0:00:56.840
<v Speaker 2>of people that could name, identify, and explain what Deep

0:00:56.880 --> 0:00:59.600
<v Speaker 2>Seek is would have been few and far apart at

0:00:59.600 --> 0:01:03.200
<v Speaker 2>myself included. I'd say potentially you included in that as well.

0:01:03.720 --> 0:01:06.240
<v Speaker 2>And now it's all anyone's talking about. It is literally

0:01:06.240 --> 0:01:11.000
<v Speaker 2>dominating every single headline across the world today. Sam, take

0:01:11.080 --> 0:01:13.600
<v Speaker 2>us back to the beginning. AI is a very very

0:01:13.600 --> 0:01:17.720
<v Speaker 2>complicated space. But what exactly is Deep Seek and why

0:01:17.760 --> 0:01:19.160
<v Speaker 2>the hell are we hearing so much about it?

0:01:19.560 --> 0:01:22.640
<v Speaker 1>So DC is a Chinese AI company that just launched

0:01:22.640 --> 0:01:25.800
<v Speaker 1>its latest AI model called R one. Now when I

0:01:25.840 --> 0:01:27.520
<v Speaker 1>say model, just think of a brain.

0:01:27.800 --> 0:01:28.400
<v Speaker 3>I can do that.

0:01:28.480 --> 0:01:30.560
<v Speaker 1>All of these AI companies are trying to work out

0:01:30.600 --> 0:01:33.760
<v Speaker 1>how to create the best AI brain. Some are better

0:01:33.800 --> 0:01:37.679
<v Speaker 1>at solving mathematical problems, others are better at expressing, reasoning,

0:01:37.840 --> 0:01:41.479
<v Speaker 1>or being emotional, trying to mimic human emotion. And of course,

0:01:41.480 --> 0:01:43.920
<v Speaker 1>the goal for every AI company is to try and

0:01:44.000 --> 0:01:46.480
<v Speaker 1>build a model or a brain that can do everything.

0:01:46.680 --> 0:01:49.200
<v Speaker 2>Okay, And so when you're talking about AI companies, an

0:01:49.200 --> 0:01:52.880
<v Speaker 2>example of that would be open Ai, Creator, Chat ChiPT exactly.

0:01:52.960 --> 0:01:55.400
<v Speaker 1>But there are many companies and there are many companies

0:01:55.520 --> 0:01:58.320
<v Speaker 1>that are in the AI space without making models, and

0:01:58.320 --> 0:02:01.360
<v Speaker 1>we'll talk about that later as well. Now, what's fascinating

0:02:01.360 --> 0:02:04.040
<v Speaker 1>about deep Seek is that they've built a model that

0:02:04.120 --> 0:02:07.760
<v Speaker 1>they say performs as well as the best American AI systems,

0:02:08.000 --> 0:02:10.160
<v Speaker 1>but at a fraction of the cost. So to put

0:02:10.200 --> 0:02:13.239
<v Speaker 1>that in context, just last week we saw major US

0:02:13.280 --> 0:02:17.520
<v Speaker 1>companies like Microsoft, Oracle and open Ai, which you just mentioned,

0:02:17.840 --> 0:02:22.800
<v Speaker 1>pledge five hundred billion US dollars to build new AI infrastructure.

0:02:23.080 --> 0:02:25.440
<v Speaker 1>Then we also had an announcement from Meta they came

0:02:25.440 --> 0:02:28.000
<v Speaker 1>out separately and said that they were committing sixty five

0:02:28.080 --> 0:02:31.680
<v Speaker 1>billion US dollars to build infrastructure. There is so much

0:02:31.800 --> 0:02:35.560
<v Speaker 1>money being spent on this technology at the moment. Then

0:02:35.560 --> 0:02:38.359
<v Speaker 1>deep Seat comes along and they say they built their

0:02:38.360 --> 0:02:42.320
<v Speaker 1>system for about six million dollars, so million versus billion.

0:02:42.880 --> 0:02:45.239
<v Speaker 1>One expert I was reading called it a joke of

0:02:45.320 --> 0:02:45.840
<v Speaker 1>a budget.

0:02:46.120 --> 0:02:49.160
<v Speaker 2>So they're clearly a lot of differences between these companies.

0:02:49.320 --> 0:02:51.160
<v Speaker 2>But all of the companies that you were just talking

0:02:51.200 --> 0:02:55.040
<v Speaker 2>about before are American companies. Deep Seek is a Chinese.

0:02:54.720 --> 0:02:57.519
<v Speaker 1>Chinese AI company that say it can do what American

0:02:57.520 --> 0:03:00.320
<v Speaker 1>companies can do for a fraction of the cost.

0:03:00.440 --> 0:03:02.720
<v Speaker 2>Okay, So then tell me more about deep Seek, like,

0:03:02.760 --> 0:03:04.280
<v Speaker 2>how is it possible that they're doing this?

0:03:04.320 --> 0:03:06.639
<v Speaker 3>Are they telling the truth? Where have they come from?

0:03:06.919 --> 0:03:10.000
<v Speaker 1>It's a really interesting story. So it started as a

0:03:10.160 --> 0:03:13.960
<v Speaker 1>side hustle project of a Chinese it's an absurd side hustle,

0:03:13.960 --> 0:03:16.840
<v Speaker 1>it's crazy of a Chinese hedge fund called high Flog.

0:03:16.680 --> 0:03:17.919
<v Speaker 3>Usually like a jewelry brand.

0:03:18.520 --> 0:03:20.960
<v Speaker 1>So these are coders who had been employed to build

0:03:21.000 --> 0:03:23.640
<v Speaker 1>systems that would trade stocks and be able to predict

0:03:24.120 --> 0:03:27.280
<v Speaker 1>for clients when to buy and sell. And they started

0:03:27.320 --> 0:03:30.320
<v Speaker 1>playing around on the side of that with different ways

0:03:30.320 --> 0:03:32.960
<v Speaker 1>to build AI products that could be used by consumers,

0:03:33.000 --> 0:03:35.560
<v Speaker 1>So by you and me. The founder then spun off

0:03:35.800 --> 0:03:38.440
<v Speaker 1>Deep Seek as a separate company to this hedge fund

0:03:38.720 --> 0:03:39.880
<v Speaker 1>only in twenty twenty three.

0:03:40.000 --> 0:03:41.440
<v Speaker 3>Oh wow, so very recent.

0:03:41.680 --> 0:03:46.200
<v Speaker 1>Yeah, and they started only recently playing around with old GPUs,

0:03:46.960 --> 0:03:49.240
<v Speaker 1>which what is one of the key tools you need

0:03:49.280 --> 0:03:52.600
<v Speaker 1>to build an AI machine that they had available to them.

0:03:52.880 --> 0:03:54.160
<v Speaker 3>What's what's GPUs?

0:03:54.240 --> 0:03:55.480
<v Speaker 1>So let's break it down. I'm going to use a

0:03:55.560 --> 0:03:59.360
<v Speaker 1>car analogy. Let's pretend that Ferrari were spending billions of

0:03:59.400 --> 0:04:02.320
<v Speaker 1>dollars a year on new car engines to try and

0:04:02.360 --> 0:04:04.560
<v Speaker 1>figure out how to win the Formula One, and then

0:04:04.600 --> 0:04:07.480
<v Speaker 1>all of a sudden, engineers at Holden, you know, a

0:04:07.520 --> 0:04:10.720
<v Speaker 1>really un sexy car company, started playing around with a

0:04:10.760 --> 0:04:14.400
<v Speaker 1>Ferrari part made fifteen years ago, and they approached the

0:04:14.520 --> 0:04:17.000
<v Speaker 1>challenge of making the fast car from a different perspective

0:04:17.200 --> 0:04:20.120
<v Speaker 1>with these old parts, and actually they've made a faster car.

0:04:20.440 --> 0:04:22.000
<v Speaker 1>That's basically what's happened here.

0:04:22.080 --> 0:04:25.239
<v Speaker 2>Okay, So what you're saying is that this Chinese company,

0:04:25.279 --> 0:04:30.960
<v Speaker 2>Deep Seek has created this basically beast of Nai machine,

0:04:31.240 --> 0:04:34.520
<v Speaker 2>are using old parts, old GPUs, which is just the

0:04:34.560 --> 0:04:37.279
<v Speaker 2>little chips yeah sure, rather than building new ones.

0:04:37.120 --> 0:04:38.679
<v Speaker 1>And getting exactly the same results.

0:04:38.880 --> 0:04:40.960
<v Speaker 3>Okay, And is that why the cost is lower?

0:04:41.080 --> 0:04:44.600
<v Speaker 1>Exactly? And so they say though that it's not actually

0:04:44.760 --> 0:04:47.560
<v Speaker 1>the new GPUs, the new chips that you need to

0:04:47.600 --> 0:04:51.719
<v Speaker 1>make really good AI, it's their innovative training techniques. So

0:04:52.160 --> 0:04:55.080
<v Speaker 1>they say they've found ways to make AI models think

0:04:55.279 --> 0:04:58.640
<v Speaker 1>more efficiently. And there was one really key part of

0:04:58.680 --> 0:05:01.360
<v Speaker 1>the research that they say makes their AI different, and

0:05:01.440 --> 0:05:05.640
<v Speaker 1>that's recognizing that AI has a heart moments. This is

0:05:05.680 --> 0:05:08.440
<v Speaker 1>actually what they say. They say that AI has a

0:05:08.560 --> 0:05:11.520
<v Speaker 1>heart moments where they start processing a trying to answer

0:05:11.560 --> 0:05:13.560
<v Speaker 1>a question you put to them, and then they realize

0:05:13.560 --> 0:05:15.680
<v Speaker 1>something that might not be on the exact path they

0:05:15.680 --> 0:05:18.440
<v Speaker 1>were following. They've trained their AI to take that route,

0:05:18.600 --> 0:05:21.080
<v Speaker 1>so they say, oh, I've actually just thought of this, this,

0:05:21.160 --> 0:05:23.880
<v Speaker 1>and this, and then go down that route. So instead

0:05:23.920 --> 0:05:26.400
<v Speaker 1>of it explicitly trying to get to the answer as

0:05:26.440 --> 0:05:29.320
<v Speaker 1>fast and direct as possible, it almost kind of goes

0:05:29.400 --> 0:05:31.480
<v Speaker 1>with the flow a bit more. And I'm really simplifying

0:05:31.480 --> 0:05:31.760
<v Speaker 1>this here.

0:05:31.760 --> 0:05:33.280
<v Speaker 3>This train is legitimately heading.

0:05:33.320 --> 0:05:37.080
<v Speaker 2>So what you're saying here is that they are suggesting

0:05:37.200 --> 0:05:40.520
<v Speaker 2>that their technology is superior because they've trained it to

0:05:40.600 --> 0:05:41.320
<v Speaker 2>think differently.

0:05:41.720 --> 0:05:45.440
<v Speaker 1>Essentially, I think that we're probably summarizing artificial intelligence. Yeah,

0:05:45.440 --> 0:05:48.320
<v Speaker 1>and I think we've probably summarized like twenty five years

0:05:48.320 --> 0:05:53.120
<v Speaker 1>of quantum physics and quantum computing into six sentences. But yes, essentially,

0:05:53.320 --> 0:05:56.000
<v Speaker 1>they've figured out how to make old technology think differently

0:05:56.040 --> 0:05:58.200
<v Speaker 1>and be as good as new technology. Okay, that's what

0:05:58.240 --> 0:06:01.240
<v Speaker 1>you need to know. Okay, Sure, And so the USAI

0:06:01.320 --> 0:06:03.960
<v Speaker 1>companies get wind of this, and they start testing this

0:06:04.080 --> 0:06:07.960
<v Speaker 1>Chinese AI yep, and they give deep Seek and open

0:06:08.000 --> 0:06:10.840
<v Speaker 1>AI's top models, so the best chatch apt models that

0:06:10.880 --> 0:06:13.680
<v Speaker 1>you can possibly get. They give the both models the

0:06:13.720 --> 0:06:18.080
<v Speaker 1>same questions, and deep Seek produces similar, if not better answers.

0:06:18.160 --> 0:06:20.239
<v Speaker 1>When was this this week? Last week?

0:06:20.520 --> 0:06:21.599
<v Speaker 3>A panic?

0:06:21.920 --> 0:06:24.640
<v Speaker 1>Exactly? Now there is an important caveat that I want

0:06:24.640 --> 0:06:28.280
<v Speaker 1>to mention here. These us AI companies have said that

0:06:28.400 --> 0:06:31.520
<v Speaker 1>whilst the deep Seek model is impressive at writing in

0:06:31.600 --> 0:06:35.400
<v Speaker 1>general problem solving, it does struggle with certain specific tasks,

0:06:35.480 --> 0:06:39.400
<v Speaker 1>so things like you know, booking and a play, Yeah,

0:06:38.160 --> 0:06:42.640
<v Speaker 1>you're you do struggle with certain specific tasks numbers. So

0:06:42.720 --> 0:06:45.040
<v Speaker 1>now there's going to be a whole independent testing phase

0:06:45.360 --> 0:06:46.880
<v Speaker 1>that will give us some clearer results.

0:06:47.000 --> 0:06:47.279
<v Speaker 3>Okay.

0:06:47.360 --> 0:06:50.039
<v Speaker 2>The reason that the AI industry basically is freaking out

0:06:50.400 --> 0:06:53.480
<v Speaker 2>is because this new players come along is doing something

0:06:53.520 --> 0:06:55.839
<v Speaker 2>for the fraction of the cost, is getting similar results,

0:06:56.200 --> 0:06:57.880
<v Speaker 2>and it's non an American company.

0:06:58.040 --> 0:07:01.599
<v Speaker 1>Yeah, okay, and it challenges this fundamental assumption that we

0:07:01.760 --> 0:07:05.400
<v Speaker 1>have about AI development, which is this conventional thinking that

0:07:05.440 --> 0:07:08.800
<v Speaker 1>you needed the most expensive, the most cutting edge computer chips,

0:07:09.000 --> 0:07:11.440
<v Speaker 1>particularly those made by one company in the video, which

0:07:11.480 --> 0:07:13.200
<v Speaker 1>is one of the most valuable companies in the world.

0:07:13.240 --> 0:07:16.080
<v Speaker 1>We've talked about it on this podcast before, to build

0:07:16.080 --> 0:07:19.200
<v Speaker 1>the best AI systems, and that's why tech companies like

0:07:19.200 --> 0:07:21.680
<v Speaker 1>Google and Meta have been spending insane amounts of money

0:07:21.960 --> 0:07:25.840
<v Speaker 1>building the infrastructure that they say they need to produce

0:07:25.880 --> 0:07:27.960
<v Speaker 1>these AI systems, and investors have gone with it.

0:07:28.040 --> 0:07:31.360
<v Speaker 2>Okay, But I guess the reason that the tech world

0:07:31.440 --> 0:07:33.840
<v Speaker 2>is in so much disarray this week is because that

0:07:33.920 --> 0:07:37.160
<v Speaker 2>type of logic might not necessarily be the only.

0:07:36.920 --> 0:07:39.760
<v Speaker 1>Way exactly, and Deep Seek says that might not be true.

0:07:39.880 --> 0:07:42.880
<v Speaker 1>And what's more, Deep Seek have released their program as

0:07:43.000 --> 0:07:46.360
<v Speaker 1>open source, and what that means is they're an open book.

0:07:46.680 --> 0:07:49.400
<v Speaker 1>They've taken the hold and around the racetrack popped open

0:07:49.440 --> 0:07:51.200
<v Speaker 1>the hood and said, whoever wants to come and have

0:07:51.240 --> 0:07:52.520
<v Speaker 1>a look as much as.

0:07:52.360 --> 0:07:54.239
<v Speaker 3>You want, Why can't they just be copied?

0:07:54.560 --> 0:07:57.800
<v Speaker 1>They say that they don't care really if they're copied.

0:07:57.920 --> 0:08:00.080
<v Speaker 1>That this technology almost think about it, like you know,

0:08:00.160 --> 0:08:01.760
<v Speaker 1>there's been a lot of examples in the last twenty

0:08:01.760 --> 0:08:03.960
<v Speaker 1>four hours about the space race. They've kind of said,

0:08:03.960 --> 0:08:06.200
<v Speaker 1>everyone's trying to get to the moon. Here's how we're

0:08:06.200 --> 0:08:09.520
<v Speaker 1>doing it, hoping that the entire industry becomes more profitable.

0:08:09.640 --> 0:08:11.840
<v Speaker 1>It's just a different way of doing business. It's very

0:08:11.880 --> 0:08:14.240
<v Speaker 1>common in the tech world to release something out to

0:08:14.360 --> 0:08:16.440
<v Speaker 1>the public. And I want to bring up one new

0:08:16.520 --> 0:08:18.880
<v Speaker 1>term here, and that's the idea of a company having

0:08:18.880 --> 0:08:22.200
<v Speaker 1>a moat, a mote e moche. So think about a

0:08:22.240 --> 0:08:25.800
<v Speaker 1>competitive advantage that surrounds a company and literally think about

0:08:25.800 --> 0:08:30.640
<v Speaker 1>a castle with emotion around it and moat. Yeah yeah,

0:08:30.680 --> 0:08:33.280
<v Speaker 1>And that's what makes a company hard to copy. So

0:08:33.480 --> 0:08:36.320
<v Speaker 1>if a company has I feel that far ahead, yeah,

0:08:36.320 --> 0:08:39.680
<v Speaker 1>but if you have a wide moat, then other companies

0:08:39.679 --> 0:08:42.559
<v Speaker 1>can't copy you quickly and easily. And so for these

0:08:42.600 --> 0:08:47.080
<v Speaker 1>AI companies, these really valuable stocks. Their moats are not

0:08:47.120 --> 0:08:49.160
<v Speaker 1>only the knowledge they have and the expertise, but just

0:08:49.160 --> 0:08:51.720
<v Speaker 1>how much cash they've got, because the common reasoning has

0:08:51.720 --> 0:08:55.080
<v Speaker 1>always been to build great AI, you need fifty sixty

0:08:55.120 --> 0:08:56.840
<v Speaker 1>seventy one hundred, five hundred billion.

0:08:57.200 --> 0:08:57.720
<v Speaker 3>I understand.

0:08:57.720 --> 0:09:01.120
<v Speaker 2>So you're saying these companies they're competitive advantages that they've

0:09:01.200 --> 0:09:05.320
<v Speaker 2>raised in dollars of money, and therefore the average Joe

0:09:05.400 --> 0:09:07.120
<v Speaker 2>on the street can't go and build the same tech

0:09:07.160 --> 0:09:09.320
<v Speaker 2>as them because they can't raise a bajillion.

0:09:08.920 --> 0:09:11.839
<v Speaker 1>Dollars until now, until now when DeepC comes out and.

0:09:11.760 --> 0:09:14.000
<v Speaker 2>Says, actually, you don't need a bajillion dollars, you only

0:09:14.000 --> 0:09:16.640
<v Speaker 2>a little bit, only six million, and you can build

0:09:16.679 --> 0:09:17.880
<v Speaker 2>something just as good.

0:09:18.040 --> 0:09:18.480
<v Speaker 3>Okay.

0:09:18.640 --> 0:09:22.280
<v Speaker 2>So what has all of this kind of knowledge and

0:09:22.400 --> 0:09:26.400
<v Speaker 2>information done to the market, because there has been quite

0:09:26.440 --> 0:09:29.080
<v Speaker 2>a significant economic kind of market impact.

0:09:29.160 --> 0:09:31.800
<v Speaker 1>Yeah, it's been really dramatic. So, as I mentioned earlier

0:09:31.800 --> 0:09:34.280
<v Speaker 1>in the video, the company that makes the computer chips

0:09:34.559 --> 0:09:37.520
<v Speaker 1>that AI companies buy to build these big AI brains,

0:09:38.000 --> 0:09:40.640
<v Speaker 1>they lost a whopping get ready, five hundred and eighty

0:09:40.720 --> 0:09:44.160
<v Speaker 1>nine billion US dollars in the market value on the

0:09:44.240 --> 0:09:47.320
<v Speaker 1>US stock market yesterday. They went down sixteen point one percent,

0:09:47.760 --> 0:09:50.520
<v Speaker 1>and the overall US market went down three point one

0:09:50.559 --> 0:09:53.240
<v Speaker 1>percent crazy. You have to remember that it's not just

0:09:53.280 --> 0:09:56.319
<v Speaker 1>companies that are directly related to AI, but there are

0:09:56.320 --> 0:09:59.080
<v Speaker 1>so many adjacent companies that have business in AI. You know,

0:09:59.120 --> 0:10:02.880
<v Speaker 1>for example, energy companies, they would be relying on the

0:10:02.880 --> 0:10:05.160
<v Speaker 1>fact that everyone's building these big data centers. They're going

0:10:05.200 --> 0:10:07.079
<v Speaker 1>to need to plug that all into a wall somewhere.

0:10:07.320 --> 0:10:10.559
<v Speaker 1>So their stocks went down. The entire market really did

0:10:10.600 --> 0:10:12.720
<v Speaker 1>freak out about the fact that there's this new way

0:10:12.720 --> 0:10:13.400
<v Speaker 1>of doing things.

0:10:13.559 --> 0:10:16.680
<v Speaker 2>And it's not just the market that responds. It is

0:10:16.720 --> 0:10:20.160
<v Speaker 2>also this kind of geopolitical angle. I've mentioned that deep

0:10:20.200 --> 0:10:23.320
<v Speaker 2>Seek is Chinese and Open Ai and the other AI

0:10:23.360 --> 0:10:27.000
<v Speaker 2>companies are American. Talk me through that geopolitical angle.

0:10:27.080 --> 0:10:29.680
<v Speaker 1>Yeah, So the US has been trying to maintain its

0:10:29.960 --> 0:10:32.800
<v Speaker 1>competitive lead in AI. They say there's a two horse

0:10:32.880 --> 0:10:35.320
<v Speaker 1>race between them and China, and the way that they're

0:10:35.320 --> 0:10:38.120
<v Speaker 1>doing that is by restricting the computer chips that are

0:10:38.200 --> 0:10:40.880
<v Speaker 1>allowed to enter China. So, to go back to our

0:10:40.920 --> 0:10:44.360
<v Speaker 1>Ferrari example, they are not letting in new models of

0:10:44.360 --> 0:10:47.160
<v Speaker 1>a Ferrari engine into China because they don't want Chinese

0:10:47.160 --> 0:10:51.479
<v Speaker 1>AI companies to rival them. But Deep Seek's success suggests

0:10:51.640 --> 0:10:55.880
<v Speaker 1>that that restriction might actually be backfiring in two possible ways.

0:10:55.960 --> 0:10:58.400
<v Speaker 1>Either the chips aren't as crucial as we all think

0:10:58.480 --> 0:11:01.719
<v Speaker 1>to building these models, or Chinese companies have worked out

0:11:01.760 --> 0:11:03.319
<v Speaker 1>ways to get around the restrictions.

0:11:03.480 --> 0:11:05.000
<v Speaker 3>Yeah. So interesting.

0:11:05.240 --> 0:11:07.200
<v Speaker 2>And when I hear about that, the one thing that

0:11:07.240 --> 0:11:10.360
<v Speaker 2>I think about is how Donald Trump would be responding

0:11:10.360 --> 0:11:12.960
<v Speaker 2>to this, because we all know that Trump's platform is

0:11:13.120 --> 0:11:17.640
<v Speaker 2>about growing specifically American companies and wealth. How has he

0:11:17.760 --> 0:11:18.559
<v Speaker 2>responded to this?

0:11:19.120 --> 0:11:21.360
<v Speaker 1>Well, this has become a core focus. I mean already

0:11:21.440 --> 0:11:23.559
<v Speaker 1>in his first week, full week of being president, he

0:11:23.559 --> 0:11:26.840
<v Speaker 1>announced a five hundred billion dollar program in collaboration with

0:11:26.920 --> 0:11:31.280
<v Speaker 1>some really big companies AI company. Yeah, and then just yesterday,

0:11:31.760 --> 0:11:33.800
<v Speaker 1>talking about Deep Seek, he said that this is a

0:11:33.800 --> 0:11:36.360
<v Speaker 1>wake up call for the AI industry and that we

0:11:36.440 --> 0:11:39.200
<v Speaker 1>need to be quote laser focused on competing to win.

0:11:39.800 --> 0:11:42.199
<v Speaker 3>Wow, it's really like Space Race.

0:11:42.480 --> 0:11:45.680
<v Speaker 1>It's very much the Space race. Trump's close advisor Elon

0:11:45.760 --> 0:11:48.160
<v Speaker 1>Musk he speaking of Space Race. Yes, he accused Deep

0:11:48.160 --> 0:11:50.120
<v Speaker 1>Seak of lying about how many chips it needed to

0:11:50.200 --> 0:11:52.520
<v Speaker 1>run system. So he essentially said, what you're doing is

0:11:52.520 --> 0:11:55.600
<v Speaker 1>simply impossible, but there is certainly a bit of panic there.

0:11:55.800 --> 0:11:58.800
<v Speaker 1>Trump said that he might use executive orders to increase

0:11:58.880 --> 0:12:01.959
<v Speaker 1>the amount of public reas sources like electricity and energy

0:12:02.000 --> 0:12:05.720
<v Speaker 1>supplies and building materials to just build these data centers

0:12:05.760 --> 0:12:08.440
<v Speaker 1>that USAI companies say they need as quick as they can.

0:12:08.800 --> 0:12:10.680
<v Speaker 2>Sam, I want to finish this chat by just bringing

0:12:10.720 --> 0:12:14.720
<v Speaker 2>it home. And I guess contextualizing this in an Australian setting,

0:12:15.240 --> 0:12:19.280
<v Speaker 2>what are the implications for Australia's tech sector? What does

0:12:19.280 --> 0:12:20.480
<v Speaker 2>this story mean for US?

0:12:20.920 --> 0:12:23.679
<v Speaker 1>Well, pretty much every country besides the US has had

0:12:23.720 --> 0:12:26.360
<v Speaker 1>the same reaction, including Australia, which is that this is

0:12:26.400 --> 0:12:29.040
<v Speaker 1>good news. We've all known that the US has a

0:12:29.160 --> 0:12:32.520
<v Speaker 1>major competitive advantage in AI, and AI is seen as

0:12:32.520 --> 0:12:35.400
<v Speaker 1>a really important part of any country's innovation strategy. So

0:12:35.760 --> 0:12:38.760
<v Speaker 1>our Industry and Science Minister Ed Husick. He said yesterday

0:12:38.760 --> 0:12:41.440
<v Speaker 1>that he expects to see more of these cheaper and

0:12:41.520 --> 0:12:44.680
<v Speaker 1>more creative AI solutions in years to come as this

0:12:44.760 --> 0:12:47.199
<v Speaker 1>becomes more mainstream, and that it kind of evens the

0:12:47.240 --> 0:12:49.800
<v Speaker 1>playing field a little bit because we don't need five

0:12:49.880 --> 0:12:53.240
<v Speaker 1>hundred billion dollars. We just need a side hustle in

0:12:54.280 --> 0:12:54.680
<v Speaker 1>our jobs.

0:12:54.720 --> 0:12:56.560
<v Speaker 3>How did an our side hustle lead to that?

0:12:56.800 --> 0:12:58.840
<v Speaker 1>I know, but I think, look, this is still a

0:12:58.840 --> 0:13:02.240
<v Speaker 1>privately held company Deep Seek, so deep Seak isn't listed anywhere.

0:13:02.400 --> 0:13:04.480
<v Speaker 1>We're going to see how the stock markets responded. The

0:13:04.480 --> 0:13:07.320
<v Speaker 1>stock markets in Australia that tech and AI companies were

0:13:07.360 --> 0:13:10.199
<v Speaker 1>hit pretty hard yesterday, but our stock market isn't as

0:13:10.240 --> 0:13:12.800
<v Speaker 1>reliant on those big tech giants as in the US.

0:13:13.640 --> 0:13:15.960
<v Speaker 1>But it's a really seismic shift in the way that

0:13:15.960 --> 0:13:18.120
<v Speaker 1>we think about AI and I'm going to have to

0:13:18.120 --> 0:13:19.600
<v Speaker 1>ask Chatcha and see what happens. Now.

0:13:20.880 --> 0:13:23.640
<v Speaker 2>Well, thank you for explaining that story. I actually feel

0:13:23.679 --> 0:13:25.319
<v Speaker 2>like I completely understand it now.

0:13:25.280 --> 0:13:25.839
<v Speaker 1>So you.

0:13:28.320 --> 0:13:28.880
<v Speaker 3>Well, there you go.

0:13:29.080 --> 0:13:31.080
<v Speaker 2>Thank you so much for joining us for another episode

0:13:31.120 --> 0:13:33.880
<v Speaker 2>of The Daily Ods and supporting our independent newsroom. We're

0:13:33.880 --> 0:13:36.200
<v Speaker 2>going to be back later today with the headlines, but

0:13:36.360 --> 0:13:37.800
<v Speaker 2>until then, have a great day.

0:13:40.640 --> 0:13:42.960
<v Speaker 1>My name is Lily Maddon and I'm a proud Arunda

0:13:43.200 --> 0:13:47.959
<v Speaker 1>Bungelung Calcottin woman from Gadigol Country. The Daily oz acknowledges

0:13:48.080 --> 0:13:50.199
<v Speaker 1>that this podcast is recorded on the lands of the

0:13:50.240 --> 0:13:53.800
<v Speaker 1>Gadigol people and pays respect to all Aboriginal and Torres

0:13:53.800 --> 0:13:56.760
<v Speaker 1>Strait Island and nations. We pay our respects to the

0:13:56.760 --> 0:13:59.480
<v Speaker 1>first peoples of these countries, both past and present,