You're listening to Asia Centric from Bloomberg Intelligence, the podcast that pulls back the curtain non global business so you can invest better across the Pacific rim. I'm Tom Corbett in Hong Kong, and.
I'm John Lee. Artificial intelligence is generating a level of investor excitement not seen since the dot com era.
Call it a feeding frenzy, call it a buying binge. The rush to board the AI bandwagon has powered America's Magnificent seven stocks to new heights, and giddy investors are rushing to get on board.
Will artificial intelligence deliver on the high or investors at risk of being left empty handed? And what about Asian tech stocks? Can they catch up to their American counterparts.
Let's bring in two Bloomberg Intelligence experts. Mandeep Seeing is Global head of Technology Research based in New York.
And Robert Lee see Asia Tech Analysts based here in Hong Kong.
Man Deep and Rob welcome to Asia Centric.
Thanks for having us on and y'are excited to be on the podcast.
Like wife Tom and John, Thanks for having us Mandy.
Let's get straight into it. Are we in an AI driven tech bubble?
Well? I think bubble is a term that gets used probably so frequently that in my lifetime I probably would have come across at least one hundred bubbles so far. So clearly, you know, there are times when things come across as fad and there isn't proper follow up in terms of monetizable opportunities. AI, especially generative AI, doesn't seem
to be that type of fad. And the reason I say that is what we have seen so far is not only you know, in video really getting that pricing power in terms of the GPUs that are being used for AI training, but also follow up application, whether it's on the software side with chatbots that can streamline back end operations, contact center stuff, as well as other type
of optimizations and copilot use for developers. So with Jenny and I, we are talking about workflows getting streamlined, and when we talk about workflows, they're across the board and we see a lot of potential over there.
And you also mentioned these AI companies are monetizing artificial intelligence. Is that a key difference between say the tech bubble in nineteen ninety nine.
I think so you are seeing a lot of upfront revenue for companies that are playing to this trend. And it is a secular tailwind in the sense that not only it's what Nvidia is doing right now, but think of the recurring revenue that these companies can generate through copilots or the subscription revenue streams that Microsoft is able to add on to its office bundle. I mean this is real. Companies are subscribing to this because they can
see knowledge workers getting a productivity boost. So I would say whether it delivers on the promise that everyone feels this technology brings to the fore I mean we ourselves have done our forecast, a ten year forecast where we think it's going to drive one point three trillion in incremental revenue by twenty thirty two. So we think this is going to grow at a kegar of forty percent
over the next ten years. But look, there will be ups and downs, as we have seen with any technology wave, and so far it seems to be delivering on the promise with revenue streams showing up in different companies that we follow.
Robert Lee, I know that you've been through the dot com bubble as with the rest of us. What similarities and differences do you see between the AI craze now and the dot com bubble back then, Uh.
Well, look at my reflection in the camera and a few more gray hairs, and I used to have I do remember the dot com bubble and I started my cell side career then. But I think the difference then was the sort of infrastructure was built out. The web was built out with little thought at that point as to how again, what's a business model, how is it going to be monetized? Whereas AI? You know, I guess to many people hadn't really thought or heard of AI
until chat GPT came along about a year ago. But AI has been around for decades in a form of machine learning as it's referred to, and so there's been a lot of research at university level and at corporate level going into AI for many decades. What's changed and what's helped to accelerate it along the curve is obviously some fine tuning of the algorithms which were previously focused more on numerical data and now they can handle words
and a language. And also the huge amounts of computing and storage power that is available today versus decades ago, and so that has enabled these sort of AI applications which have been there for many decades, but perhaps below the horizon, unknown and unseen so many people. It's allowed a proliferation of those applications and opening up substantial new markets and opportunities. In arguably what's going to be not only a multi year secular trend, but a multi decade
secular trend. So it's a very very exciting place to be and very very different to what things look like in the late nineteen nineties.
Manleep seeing does it surprise you how quickly AI has erupted under the investment landscape.
Not really if you follow the trajectory of evolution, and as Rob alluded to, you know, machine learning has been around for a number of years. The whole large anguage model training is about ingesting large amounts of data, using the compute capacity and really putting guardrails around the technology to the point that it's not only productive, but uh, you know, you're not at the risk of it being misused.
I think that is one of the biggest concerns that I see, is the potential for this to be misused by bad actors, and I think AI always had that promise. Clearly, if we have made strides in terms of you know, deploying large anguage models at scale and to be used in productivity apps. But the key is how do you ensure safety and you make sure that the guardrails are there for all kinds of use cases. And to me, that still remains to be proven, Man Dave.
Nvidia is probably the most important stock right now and everyone eagerly awaits when it reports its quarterly earnings. Is the bullmarket totally relying on Nvida continuing to beat estimates or is there any other metrics we should look at.
When you look at you know, team stock like Nvidia is it's supposter child for generative AI and for a good reason. And I think when you look at how they kind of became the name to resemble this wave of gen AI, it's because they had these chips that they had developed over the last thirty years around gaming and parallel processing, which was a different type of architecture when you compare it to the traditional CPU architecture that is used for pretty much all kinds of processing up
until now. So this concept of accelerators and parallel processing is new when you think about the data centers, and the reason why it's taken off in such a big way is because when you're dealing with large aguage models. We're talking about a scale that we haven't used for processing before. Right now, you're seeing supply constraints. I mean, think of how fab capacity increases over time. It normally increases you know, mid single digit, because that's how the
KAPEX investments used to be. How we are talking about insatiable demand for video chips and that to leading note chips. We're talking about three to five nanometer fabs that are needed for producing these accelerator chips and we don't have that capacity. Combine that with you know, COOS packaging that's needed for these chips. So clearly the supply constraints are
what's driving the asps. But I won't pay too much attention to asps even though that's what's driving the high growth rates, the fifty percent plus kegres that we've seen with Nvidia. It all comes down to what is the addressable market for data centers over the next ten years, and based on our one point three trillion number, half of that is hardware needed for training and inferencing. So clearly there is a long runway when it comes to hardware.
Man leeps saying how sustainable is Navidio's competitive advantage.
Do you think now, look at the biggest buyers of Nvidia's chips, it's your hyperscaler companies, the companies that deploy them on cloud to be used for widespread consumption. Anytime you see such big concentration where your customers can become
potential competitors, which is the case with Nvidia's buyers. Here, Amazon, Google, Microsoft, Meta, all of them have said they are developing their own chips because they think vertical integration would give them an edge over time, and we've seen that with the Apple model. I mean, think of why Apple's smartphone is so successful. It's a vertical integration from chips all the way to software and operating system. So clearly there is a risk.
But right now, by the time these companies develop a version of their accelerator, Nvidia would be at least four or five generations ahead in terms of the performance. So that's why you know every three to four months, these companies, especially Nvidia NAMD, they will keep progressing to the point their processor, their accelerator performance will be ahead of what the hyperscillers are developing. And I think that kind of edge will remain.
For a while Rob Lee, did you want to weigh in?
I agree with that. I mean, the semiconductor is at the lynchpin of the global supply constraints at the moment within the AI sector, you know, particularly coming from TISMC, given the supply bottleneck that exists within that business, because TSMC is the only semi conductor foundry globally with true deep experience in leading edge technology derived on production equipment
from ASML. So it's those two key companies which are at the center of this supply bottleneck which is benefiting in video at the moment and obviously causing this constraining factor.
Is clearly benefiting in video, but also to the detriment of the Chinese given the export constraints, can't access leading edge technologies, and also given the supply constraints on the semiconductor founder of fabrication side, they also can't acts leading edge semiconductor equipment in order to fabricate their own chips. So I think I just at those two points in there before.
We get onto like Asia Tech. But Mande, but I wanted to just broaden the discussion to like the stocks outside in VideA, How did you grade the Magnificent seven, Like, what scorecard would you give them to be an ABAC in terms of the different companies.
Well so, my lens right now in terms of evaluating Magnificent seven is around which out of the magnificent seven have their own foundational models and based on that, Google, Meta and Microsoft because of their early Open AI partnership, stand out to me as companies that have their own foundational models natively deployed on their cloud and that gives them an edge because we are talking about generator AI.
Amazon is the largest cloud player. Clearly they have a lot of potential to bene fit from the trend, but what they haven't figured out so far is their foundational model and what are they going to standardize on. Same thing with Apple. I mean, clearly Apple has the install based when it comes to you know, on device AI and inferencing, but we don't know much about their GENI strategy. And Nvidia clearly is the biggest beneficiary of the trend, so they get very high probably a plus so far.
So that leaves Tesla. Tesla, I think is an interesting play a derivative of their early strategy around the FSD software which wasn't built on generative AI per se, but they have changed the latest version of their software to leverage generative AI, so clearly they have a headstart when it comes to generative AI on the automotive side. But I think if you use that framework, companies with their own foundational models will have a persistent advantage than the ones that don't.
Our guests are man Deep Sing and Rob Lee, both tech analysts with Bloomberg Intelligence. Mandeep and Rob Let's talk about Taiwan Semiconductor. Some in the West refer to it as the most important company, the most important manufacturer that they've never heard of. Sixty to seventy percent of the global chip market is what it commands, talk about its positioning in the broader tech AI matrix and why should investors care?
I mean, look, and Rob probably has better visibility on what TSMC is doing in the region. But from a Western perspective, there is this reliance on TSMC when it comes to all the fabulous chip makers, including Nvidia that we wave about so much. At the end of the day, they rely on TSMC to make those GPUs that there's so much demand for. So clearly that is a choke
point when you think about deploying generative AI. But I have no doubt, and you know, we just published two deep dives supply chain and semicap equipment here at BI and part of the conclusion was that the advanced chief manufacturing capacity will diversify over the next ten years, and that's where you will see companies like TSMC as well as other big fab makers actually invest their CAPEX dollars in terms of expanding their fabs outside the East Asia region.
Robbie, you have some interesting perspectives about TSMC, including how it's affected Taiwan's exports over the past decades.
Yes, well, the I wish I had a very precise state for you here, but at some point in the nineteen sixties, I believe Taiwan's number one export was actually sugarcane,
so it was largely an agricultural based economy. And although that's sixty plus years ago, through significant government support, through cooperation with the universities in pushing people or encouraging people to study STEM subjects and engineering, the transformation of the Taiwanese economies phenomenal used to be referred to as one of the Asian tigers, if you remember. So, going back to what Mandeep was saying as well, and just a
little bit of historical context. Again, it's twenty years or so ago. At the beginning of my investing career, I think the view within the semiconductor industry was the real value is in the design companies like Armholdings which were still around then, and companies designing their chips. That's where
the value was added and was captured. The assembly. The fabrication was important, but to a lesser extent, but as a result of something referred to as Moore's law, as the complexity of cease chips has increased to the drum beat which defines to grow for it within the industry, with transistor densities doubling every eighteenth to twenty four months. Sem conductors these days are fabricated at the atomic scale, so they're incredibly technically complex and it's incredibly difficult to
fabricate at such a high level of miniaturization. And TSMC really is the only company on the planet which has really kept a commanding lead in fact enhanced and built its commanding lead over the competition over that period. So we've seen a transfer of value from the design, which is still important that increasingly to the fabrication or manufacturing side, and that is why TSMC on a global scale is
so important. Obviously with an interesting overlay of geopolitics and everything else, which makes the story both more interesting and more complex to analyze. But TSMC is really the key lynchpin of the global technology sector at the moment, I would say, and.
Robert, if you look at the share prices of Asian tech like TSMC definitely appears to have benefited from the AI trend, but some other stocks in your coverage seems to have really missed the AI hyph. I'm talking about the big Chinese tech companies Ali and tenset. Now, prior to recording this podcast, I did look at the Bloomberg terminal and if you roll back the clock three years ago, Ali, Barber and Tencent in terms of market cap were actually
comparable to the Magnificent Seven. If you look at the average market cap of the Magnificent seven now, it's almost ten times the size of Ali Barber. What went wrong with China tech, especially as it relates to AI, right, how.
Long have we got?
Is?
You don't answer for that? Well, clearly the China tech sectors being caught up with a lot of concerns on the regulatory front, a lot of domestic regulation with Chinese government want to, you know, to level up the playing field and with a lot of I think very well fought through regulation to protect the more vulnerable in society from the overarching power of these two Internet giants. That's one issue that's impacted their earnings outlook and clearly their
share price. But on the AI side, I think whilst China has some very interesting technology, on the software side, there are two big differences with the US market. The first is it's a far more competitive space than in the US. So for example, when Chat GPT was announced roughly a year ago, you hadn't heard of any Chinese large language models. At that point there were probably were a few in development behind closed doors. As of October
last year, there are two hundred and thirty eight. So this is an immensely competitive field in China, and these companies have come from a relative standing start to develop a large number of models, which also tells you that the relative technical barriers to entry are quite low, certainly within China. So again that has implications for the longer term growth potential and margin potential of these companies, and
again I think that's another thing that's factored in. But I think the most important thing which is going to increasingly forestall their growth curve is the access to the core infrastructure in which these large language models are both trained and then the inferencing happens, and that is largely
US derived technology. So whilst the chips may be fabricated in Taiwan, they are in Nvidia design chips or AMD design chips, and due to the export restrictions, it's increasingly difficult for Chinese companies to get hold of those, which makes them evra reliant on lower generation chips which are slower or more energy intensive, you know, their inferior in many respects, or it makes them more reliant on homegrown alternatives, which again are one to two generations behind, and they're
also in short supply. So if you can't get the infrastructure to run these models off, then clearly that will impact your growth curve over the next few years. So these are some of the factors that have impacted the performance of these stocks. It's more complex than that, but maybe you can do another podcast and that another day.
Robert, Have you even checked out any of them The mainland China apps relating to AI of course, yeah, and how do they compared to chapt This.
Is purely anecdotal, But myself and a colleague this is into last year when ernie bot, which is Baidu's Chat GPT like janitor of a tool, was first launched. We compared that and you know, just to put a few questions running through GPT three and a half and ernie bot, and I think by dou is perfectly adept at answering any questions relating to China, to Chinese history, to Chinese culture. No major issues however, when it comes to answering questions
on things that lie outside China. So for example, we asked it some questions on Bill Gates and just tell us a bit about him. You know, it gave two three lines, very simplistic lines without a lot of detail, comparing contrast to Chat GPT, giving significantly more information with a lot more opinion and a lot more depth and value. Add Now, why is that? There are two reasons? Again?
I think one is the access to training data. Something like ninety percent of the world's worldwide Web content is in English and roughly ten percent is in Chinese, and I think the availability of training data is key to the effectiveness of these models. So the breadth and depth of training data available to Bido's early bot is insignificant compared to chat GPT, and that is reflected in their
relative performance. And obviously, the Chinese data that anybot is trained on is largely China specific, which accounts for its superior results. The second issue, just briefly is obviously on the censorship rules. So obviously, if you asking anybot on anything that's politically sensitive, you're going to get a curtailed answer. Let's just leave that answer at that.
Mandy, look into your crystal ball. Two years, five years, maybe ten years down the road. Where is this taking us? Where is AI going to lead us? Is it going to be to a good place or a not so good place?
I think the answer there is simple. It definitely will lead us to a much better place in terms of the efficiencies and the productivity that we think generative AI promises. And the reason I feel fairly confident on that front is we have already seen the proof points when it comes to all the contextual knowledge that's embedded in these models. I mean, you're talking about intelligence systems that can understand human language and really never get tired, and you know
it can respond in a very intelligent way. So when you apply that framework to different facets of our life, the world will be a much better place. The only thing I would put as a caveat is the guardrails need to be in place, because this can potentially be misused. All these foundational models are trained on a lot of bad things as well, which are embedded in the model. That's why the guardrails are very important.
Rob Lee closing thoughts, Okay.
Yeah, I don't disagree with any of that. I think if you look at who are the interested parties within the world of AI, there are the makers of the infrastructure, so clearly that's the in videos, TSMC's armholdings, etc. And they're clearly profiting from it at the moment as the
infrastructure is built out. There are then as sort of mandeeps referred to the end users of AI, and I think there have been plenty of examples where the utilizing AI can drive significant cost and productivity savings and we're only on the cusp of that. So the potential benefits in the long term through the use of AI, you know, potentially immense. Then there are the software companies in the middle again, and I'm referring to the Chinese companies here.
On one hand, they are having to spend significant amounts on building out the infrastructure, so it's costing them tens of billions of US dollars at a time when they're
monetarzation model is at a very immature stage. So I think, again tying back to John's earlier question, I think that's one reason why we're seeing on the performance in the China AI stocks at the moment, because they're caught in the middle anyway, and until they find a better way of actually driving into profit and driving some learnings from this, I don't see that we're going to get any short term change there.
Our guests have been Man Deep Seeing, Global head of Technology Research based in New York and Bloomberg Intelligence, and Robert Lee, Senior Asia tech analyst based here in Hong Kong. It's been a fascinating discussion on AI, the risk, the opportunity, and the business involved implications for Asia and around the world. Man Deep and Rob has been great having you.
Great Thanks for having us John and Tom, and yeah, look forward to doing this again in the future.
Absolutely, I completely echo that it's been good fun. Let's do it again sometimes.
I'm Tom Corbett in Hong Kong and I'm John Lee.
This podcast was produced by Clara Chen and you've been listening to the Asia Centric podcast