- 8 months ago
- #woosays
This episode of #WooSays looks at how Artificial Intelligence has become a key frontier in the global contest for power. Professor Emeritus Datuk Woo Wing Thye and Melisa Idris explore how AI is transforming the balance of power, to the growing divide between nations that build AI and those that must live with it. All this and more, in this week’s #WooSays.
Category
🗞
NewsTranscript
00:00Hi folks, welcome to another episode of WUSAZ. This is where we get real about US, China and
00:17everywhere and everything in between. It's two superpowers, but really one global story.
00:23This week on WUSAZ, we turn to one of the defining power struggles of our time, artificial
00:31intelligence. With the US and China accelerating their rivalry on everything from chips to data
00:38to regulation, is AI the new frontier for geopolitical competition? That's what we want
00:47to explore. And joining me as always is Professor Wu Wingtai, who will help us unpack what this
00:54means for the world, what this AI race will mean and how countries should navigate between
01:00the tech superpowers. So joining me now is Prof Wu. Prof Wu, let's talk a little bit about
01:07this. So what's interesting about how AI has unfolded in our time is it seems that there's
01:15notes of history in it. So it brings us back to when there was another big boom of technology.
01:22Every time the world goes through these technological revolutions, everything changes. Global dynamics,
01:32the global economy. So talk to me a little bit about how the current AI revolution maybe brings
01:40us or there are lessons from say the industrial revolution from the past.
01:46Well, the industrial revolution certainly brought about a dramatic reconfiguration of human society.
01:54We move from an urban society to a rural society. We move from a rural society to an urban society,
02:04largely because the steam engine enabled manufacturing, which requires people to be in much closer quarters
02:14with each other than before. And it certainly brought about us great prosperity. But on the journey
02:24from the rural society to the urban society, we did go through a very painful period. A very painful period in terms of
02:35the jobs of the past basically disappeared, but there was new jobs appeared. So there was a reshaping
02:46of the distribution of the distribution of the distribution of jobs from people making horsewhips to people making
02:56leather belts and screws and nuts and bolts. What is happening is there will no doubt be a reshaping of the workplace.
03:11But the part that would hit us immediately would be the loss of jobs. In other words, the number of jobs that will be lost
03:24will be the greater than the number of jobs that are created immediately. In the case of the Industrial Revolution,
03:34that in England, there were a group of people called Luddites, because they claim to be led by Mr. Ludd, Ludd, Ludd, Luddites
03:46were displaced farm workers and factory workers who went around destroying new machines. So the first prediction of the coming of AI in the production process
04:03is that there would be layoffs of workers, not only in manufacturing, but also in the white collar level.
04:16Elon Musk has said, sorry, Elon Musk has said that the entry level white collar jobs would basically disappear.
04:28AI, of course, some people, of course, would think that not all will disappear because if you ever had called Malaysia Airlines recently, you will get a chatbot.
04:41And thank God, when you insist on talking to a human, they still give you a human. But I think there will be improvements to the chatbots such that many of our questions could be answered satisfactorily,
04:57which means that even jobs like call centers would be eliminated. So the first thing is, many people would say, isn't it great? We'll have so much leisure.
05:10Now, yes, we'll have so much leisure. But the question is, suppose, how are we going to buy the goods that are produced? If there are, because clearly you cannot just have production without people buying.
05:28So that is the first question, how are we going to distribute the big increases in output from this new technology?
05:40Elon Musk, I cannot help but to cite him, largely because he has been at the forefront of pushing for it.
05:49And he basically agrees that there will be a big loss in jobs. So he has said that there should be what we call universal basic income.
06:00In other words, we are all guaranteed a minimum level of income.
06:06It's almost like the day you are born, the state gives you three robots to support you for the rest of your life.
06:14Sorry, Elon Musk said that.
06:16Elon Musk is universal basic income.
06:20Wow.
06:21And my little bit of amendment of that is the state gives you two robots to take care of you.
06:30Okay, so if that's the case, then we want to look at the distribution of how to distribute the output that will come from the development or the growth of artificial intelligence.
06:44I do wonder how this will change the landscape.
06:48Because, again, with any kind of redistribution of growth, inequality can happen.
06:56It might not necessarily be a fair and equal distribution of the benefits of AI.
07:05How do you see this?
07:06Do you worry about an AI divide where AI exacerbates global inequalities?
07:15Well, it would exacerbate inequality even within the United States, the countries of origin of AI.
07:24If you were to look at new technology that comes about, they first have monopoly power over the ownership of the assets.
07:34And that certainly creates a tremendous amount of income.
07:38If you remember, socialism as a movement came during the Industrial Revolution because some people really got filthy rich while people were just losing their jobs with the new machinery.
07:54So the history of the world is that new technology would make redistribution of income to be more concentrated than before.
08:09Then society would have to undertake corrective actions.
08:13For example, transfer scheme.
08:16So the basic universal basic income that Elon Musk has talked about is a form of, you could say, welfarism that smooths over the pain of loss of jobs.
08:33But yes, it is likely that there will be a huge increase in inequality before it gets better.
08:44And that increase in equality comes with massive loss of jobs.
08:50So we have to think of the design of social safety nets.
08:57Basically, how would Malaysia Madani be reconfigured in order to ensure inclusiveness among the rest of the people in the country?
09:12Right.
09:13And yes, it will exacerbate the gap between countries.
09:18Because the poorest countries, but the fact is that the cost of the new technology will go down over time.
09:31That is certainly the case anytime when there's a new widespread technology that comes about.
09:42There will be this great enthusiasm, 100 firms will appear and they will all invest like mad in the new technology.
09:54And that's when you get your bubble.
09:57And then there will be a shakeout.
09:59Then at the end of it, you might have three or four companies that dominate the industry.
10:06Do you see what's happening currently with all this talk about an AI bubble and how it's on the precipice or the verge of bursting?
10:18Do you see any similarities to the dot-com bubble of the late 90s?
10:23Is that similar?
10:25I mean, you know, there was so much talk about the dot-com bubble bursting and we still have the internet today.
10:31We still have these companies from the 90s today and they're still doing well.
10:36Do you see some similarities with what's being talked about with this AI bubble?
10:40Oh, the bubbles will be inevitable.
10:43And that will be the cost of the redistribution of wealth among the rich people.
10:51Many of them would disappear if they do not get out of the business in time.
10:56It's just like China once had over 200 EV companies and that number is shrinking.
11:08And these are all not just state supported, you know, most of them are private.
11:12People get into it because if you can last, outlast the competition, there's a lot of money to be made.
11:21And in this case, the game is so new.
11:27And so we see a great proliferation of data centers because that is the first thing that is needed in the digital transformation.
11:38Okay. Well, since you brought up China, let's take a look at the major AI players.
11:45Of course, China really is emerging as a major AI adopter, especially in manufacturing.
11:53I think FinTech also is another area that they're really, really leading the pack on.
11:58How would you say China's approach to AI is different from other countries compared to the US?
12:08So let's take the two global superpowers here.
12:11Compare how each superpower is approaching artificial intelligence.
12:17Well, you could see that both countries recognize that there is a technological revolution that would greatly transform the production landscape.
12:34But they also see that it is not just economic competitiveness that could be changed.
12:43It would be the balance of power.
12:47Who would have the more powerful AI weapon than the other person?
12:52So we are on the cusp of the world having to make a decision.
13:01It's like after we found the splitting of the atom, how much of the splitting of the atom will be used for energy, you know, to support the industry, or we use it to lock, to destroy the other party.
13:16So peace and getting along has become even more important than before.
13:26Because AI is, you could see this in drones.
13:34The first drones that were used very successfully in reversing the fortune of a battle was in Ethiopia.
13:44A few years ago, when Tigray and the federal government within Ethiopia had a fight, the Tigrayans were within mouths of Addis Ababa.
13:57And what happened?
13:58Then they were all pushed back because the federal government bought drones from Turkey.
14:05And that was able to destroy the supply lines of Tigrayans and pushed them back into Tigray.
14:16Now, the important thing to realize is that we have got a lot more sophisticated than that now.
14:23In the old days, the drone actually has to be driven, someone controls it remotely, and someone has to press a button to fire the rocket.
14:38But now, in the Ukraine war, we have autonomous drones, drones that take off, and they analyze the situation on the ground, and it fires at its, according to its programming.
15:00So there's no human holding the trigger anymore.
15:06Now, the question is, what if we reach the stage where, if we happen to, the academic come a time where AI becomes smarter than people.
15:22And if one of them would have turned rogue because of misprogramming or maybe lightning struck it and the circuit went from good guy to bad guy, how do we turn off the rogue drone?
15:44Because after all, machinery going wrong happens all the time because we have robots in production, in use, and sometimes some robots start throwing the product that it was making because there was some electrical fault in the motherboard and the drone just go mad.
16:05And of course they switch it off.
16:06Now, the question is, a super smart robot may be able to resist your ability to shut it down.
16:18And in the design of the new weapon system, there may not be a shutdown button.
16:24Largely this is to prevent the other side from hijacking your drone and turning it against you or switching it off.
16:37So, we as a group have to decide on the peaceful use of AI.
16:48Yeah.
16:49That is very important.
16:50Yeah.
16:51I mean, that is such a really good point because as AI becomes more central to not just economic power but military power, I think it's important for countries, all countries, to think about how they are shaping global AI governance standards, right?
17:09And it's not just rules that are written by the superpowers.
17:14I want to come back to that a bit later, but I just want to follow up with you on how you see the US.
17:21So, we talked a little bit about how this is maybe being adopted globally and how it could be adopted.
17:28But what are your thoughts on how the US is approaching AI?
17:33Is it similar to that of China? Are they taking a more progressive or forward-looking approach to AI, especially under the Trump administration?
17:46Well, during the Biden administration, you could certainly hear a lot of words of caution about the need for careful regulation of AI.
18:00And in fact, the person who's called the godfather of AI, Geoffrey Hinton, he used to work for Google.
18:10He was the head of the Google team on AI.
18:13He resigned so that he could speak up against the way that AI was being developed at Google, largely because he thought that there was not enough safety being built into the protocol of releasing AI.
18:37So, because of the efforts of people like Geoffrey Hinton, there was a lot of responsiveness on the federal government.
18:53But in the case of Donald Trump, he has certainly been very gung-ho about it, having extreme faith in that the private sector would regulate itself.
19:10And of course, that was our attitude towards the appearance of new financial instruments that finally led to the meltdown in 2008 when mortgage-backed securities caused the meltdown in Wall Street.
19:30So, we need more regulation, and I hope that we do not introduce them only after a bad disaster.
19:42Right.
19:43In the case of China, there is more control over AI, largely because the government recognizes that AI is an instrument that could undermine its control of society.
20:01For example, the internet, you just think how much easier people could now talk to each other, the pictures and so forth.
20:10And the worst part is that AI has become so good, we can produce videos that are just fake.
20:24Yeah.
20:25Yeah.
20:26And AI, unless we are very careful on how it's regulated, it will undermine any government in the world.
20:38Right.
20:39Yeah.
20:40By completely destroying trust in news, for example.
20:44Yeah.
20:45As a news media practitioner, I am terrified.
20:49Have you seen how good those new AI videos are?
20:53It is virtually indistinguishable from photographic evidence.
21:00I mean, is the age of video evidence now no longer applicable because we just cannot trust what we see on the internet?
21:08It's so good, those AI videos.
21:11It is so realistic.
21:13Like for example, a friend of mine, Jeffrey Sachs, told me that around 10 fake videos of him appeared every day.
21:24But there's nothing he can do to stop it.
21:28Because YouTube has protected itself with laws that it bought in Washington DC, that they are not responsible for the content that is shown on their website.
21:43So it says, I'm just a platform.
21:46I didn't put it there.
21:48And I think no responsibility for someone making your, you know, making you sound stupid.
21:55So that is, it requires actions that are quite comprehensive across the board.
22:11And of course, regulation and regulations always lag behind innovation.
22:18I feel for him for having all those videos.
22:21I mean, I have some out there as well.
22:23And I just want to put it out on the record that I will never promote any dodgy real estate investments.
22:28If you see a video of me doing that, that is definitely not me.
22:33It is AI.
22:34Can I, before we move into what this means for the rest of the world, can I ask you one more question about China and the US in terms of the talent that it takes to continue this AI revolution?
22:49Because it requires, and I don't know whether China is leading in this, in terms of AI talent or the necessary education system, the necessary brains that go into developing new AI or taking AI forward.
23:10Is the US able to catch up with China in this race?
23:14And correct me if I've got that wrong, whether my assumptions of China leading this AI talent race is incorrect.
23:23Well, it is hard to measure who is leading, because what's the metrics?
23:28Because you could see that 60% of what we call to be AI patents are held by China.
23:37Now, the question is, are they the key patents?
23:41Because you can always, you know, patent your invention.
23:48But the question is, is that invention finally become the industry standard or not?
23:54So even though China is ahead in the numbers, what it would be hard to say that it controls the most important technology.
24:07Because right now, it is generally agreed that Silicon Valley is ahead in this.
24:15And for example, you could see the latest chips are still designed, the most advanced chips are still being designed in the United States.
24:27Albeit that the Chinese have claimed that they have now come up with chips that does quantum computing, which is several levels of higher speed.
24:38Now, suppose you're able to design a much more powerful chip, the next step is, can you translate it into a chip?
24:47Can you make it?
24:48In other words, can you, because the more powerful a chip is, the more circuits there are within the chip.
24:55And the more circuits there are, the hotter the chip is, and if it becomes too hot, it stops working.
25:01So you got to make it very thin so that it can be cooled down quickly.
25:08And one thing we do know is the Chinese cannot make chips as thin as the Americans have been able to do.
25:18The ASML company, which is the company that makes the machine that makes the small, thin chips.
25:28Although that may not necessarily be able to hold the Chinese back because you use a less efficient chip, but then you, you, you may be 10 seconds behind the other side in generating an answer.
25:44But you don't necessarily not get the correct answer.
25:51But more importantly, most of inventions, that takes place when you know that the other side has succeeded in doing a particular task.
26:05Once you know that it is possible to do something, you are very focused in your search.
26:14And so people do catch up.
26:16Right.
26:17And then, and the, the, the way it happens is, unless you don't sell any of your latest products,
26:26once you sell one unit of your latest product, the other side, when you get hold of one, just unpack it and see how you do something.
26:37And they, and they figure out how to make the same thing.
26:42So, uh, you could say it is good that we are able to learn from each other.
26:47And, uh, that is what economics, uh, values as competition.
26:53Right.
26:54The competition is good.
26:55So we've talked about us and China and how they're adopting AI, but I do want to ask for, maybe if we brought it home for a country like Malaysia, um,
27:05we're neither an AI, you know, frontier developer, we're not really a lay adopter, we're, we're in the game.
27:14But, um, what do you think countries who are in the middle of these big superpowers, how should they think about AI?
27:26And, um, especially when it, when we know AI will impact the labour market, will impact the, um, socioeconomic, um,
27:38will impact everything, consumption patterns.
27:40Yes, definitely.
27:41So how do you think we should, countries like Malaysia should, um, approach this?
27:47Well, one thing for sure is a society that is prepared in the following ways.
27:58One is to be able to participate in industries that use AI intensively.
28:08So the training of people, that technical skill is important.
28:14But what is also important, the only way we can really get ahead is that you've got to be more creative than the average programmer.
28:24In other words, uh, I cannot out memorize what ChatGPT already know, but I can think of clever ways to do things than what ChatGPT has collected from other people so far.
28:41So, in a way, education has to take the stress on kind of, uh, getting people, the best of its population to develop the critical thinking skill, enhance creativity much more.
29:00And I, I, I myself think that, uh, distribution of intelligence is the same across every, uh, label, uh, racial group, label, uh, income group, or whatever.
29:15Uh, but what, uh, but what in the end that matters is that you are able to have, uh, synergy.
29:27In other words, the cleverest 10% in Malaysia have to be in big constant touch with the cleverest 10% in the US, the cleverest 10% in China.
29:39Uh, so that we are able to be up to date in what they are up to.
29:46So, uh, uh, the advantage of a country like China, when you ask whether it's a hate or not, is that the absolute number of the 10% is much bigger than the 10% in the US.
29:58And one thing we do know about analytical intellectual work is that the more smart people together, they form a critical mass such that what they produce is greater than what the sum of individually, individual contribution.
30:21So, uh, we hope that common global prosperity is make more possible because we are able to keep all the 10, the, the top 15% of every country very connected to each other.
30:41And what we need to teach them is you've got to take care of the other 85% too.
30:46But the truth is the other 85% would be able to, uh, will have to do other things.
30:53And unfortunately, they, they, they will be the one that we have to focus the universal basic income coverage for.
31:01Right. Right.
31:03Well, I guess that's all the time we have for this episode of Wu says, um, Prof Wu, thank you so much for getting real about the US and China and everything and everywhere in between,
31:15particularly about AI in this episode.
31:18So that's it for this episode of Wu says.
31:20I'm Melissa Idris and I will see you in the next episode till then.
31:24Bye bye.
Comments