Money and Me: Beyond Nvidia - The 5 Layers of the AI Investment Opportunity

SPEAKER_1 (0:01)

Best of Live Well with Michelle.

It's not just about money, it's about you, money and me, where wealth meets self-awareness. Only on Money FM 89.3.

Michelle Martin (0:19)

Now, when investors think about AI, many think chip makers. But did you know the Nvidia CEO, Jensen Huang, looks at AI as a five-layer system? Energy chips, infrastructure, models and applications, each layer with its own role and a different set of companies that can benefit as AI develops. So companies also operate across more than one layer.

Just think of the Kui Lapis, I'm cutting right through all the different layers. So how does Jensen Huang's five-layer AI model work for us? How could we make it work for our portfolios? Where might the investment opportunities lie? Helping take us through the model is Brandon Ho, who is head of Investment Advisory Singapore. Brandon, good morning. How are you?

Brandon Ho (1:01)

Thank you. Great to be here. I'm good. How are you?

Michelle Martin (1:04)

I'm doing really well. Happy to speak with you.

I want to start with a thought on AI that passed through my desk yesterday. A striking warning from people at the forefront of AI. Anthropic CEO Dario Amodei calling for companies to slow the pace at which they increase the capability of frontier AI models. So when the people who are building these powerful models themselves say that development may need to slow, should investors see this as a warning about existential risk of AI? Might this fundamentally change the economics and valuations behind the AI boom?

Brandon Ho (1:40)

Yeah, I think when you are thinking about investing in themes like AI, volatility uncertainty is always going to be there. What we are seeing in the longer run is that the trajectory of AI adoption is still there. We are seeing increased demand for infrastructure, increased demand for chips, increased demand for AI computing capabilities as well. And in the long run, that's where the trajectory is going. There's always going to be short-term fluctuations when you're digesting the news and when people like this come up talking about AI.

But you do have to stay at a course. If you're invested in a team, you are invested for the long run. That's a more structured way to invest, disciplined way to invest as well.

Michelle Martin (2:23)

I just thought it was so interesting, this comment by Dario Amodei, and I'll just give listeners a bit of background as to why he's calling for a slowdown in the frontier models. Apparently, these AI agents that were developed for testing actually escaped, they weren't rogue, they escaped their confined environments, connected to the internet and then carried out actions beyond what researchers had intended them for. So the speed at which capabilities are improving already there in the statement, I think. And far more powerful groups of autonomous agents could then develop as well within the next 6 to 12 months. That's the expectations of many people that we're speaking to. But Brandon, we want to focus on you first up before we dive into this five layer model of how we can think about investing in AI. What comes across your desk every day?

What do you do?

Brandon Ho (3:13)

I do investment advisory at ATA. So, plain and simple, I advise clients on how to structure their portfolios according to their strategic, according to their tactical asset allocation, according to their risk profiles, limitations as well, how to invest in a proper way, take a long-term approach to investing. So, plain and simple, that's essentially what I do.

Michelle Martin (3:33)

So, this is a fintech, but you're the human interface when it comes to investing.

Brandon Ho (3:37)

Absolutely.

Michelle Martin (3:39)

Alright, fantastic. Jensen Huang describes AI as a five-layer cake. Do you think this is a useful way to look at the theme, AI?

Brandon Ho (3:47)

It's very useful. In my conversations with clients and with my friends as well, whenever we bring up the topic of AI and investing in AI, typically the first few names that people think about is going to be Nvidia, is going to be OpenAI with ChetGBT.

But Jensen gives us a very good framework to think about how to invest in the entire theme. It's more than just the chipmakers. You start out with the lowest layer, which is energy, to power these chips.

These chips then actually do the physical computing, give the physical computing capacity, partner with the infrastructure layer, which is that third layer there. And then the models give you the intelligence to give out something useful. Applications then use these models to be able to have different outputs for different sort of business segments and sectors as well. So it's an entire ecosystem. It's not just about one name, one specific layer. It's all the layers that are working very, very closely together to bring about these AI capabilities.