Nick Griffin: This is just the beginning of the AI runway

So he sat down all these big executives and said, "We're going all in on AI." They increased their investment in OpenAI from US$1 billion to US$10 billion. They launched their copilot products across every single product that Microsoft produces. So whether it's Dynamics or Office, PowerPoint, Excel, Outlook, etc., we'll have a copilot attached to them probably by the end of this calendar year. And he's essentially harnessed the entire company towards this one opportunity. And the reason why he's doing it is because he doesn't want to miss it like they missed the mobile bet.

And what we now see is Microsoft's CapEx budget has gone up by 50%. They're going all in and I think it's 100% the right thing to do. It's 100% the right thing to do because they've got themselves a good lead here and you're now seeing the entire tech industry try to follow them. And I suppose the big question is, are we going to use all these products when they come out when Microsoft launches them later this year?

And I think that's as Jensen Huang, the President of NVIDIA, would put it, that's your iPhone moment for AI. What's essentially happened here is you now have applications that are effectively infinite. And the reason why he calls it an "iPhone moment" is if you think about it, when the iPhone came along, it was great. The internet was on your phone, but it was really the Apple ecosystem that made the iPhone work. It could track your health, it could create a map, it could play a podcast. So you should think about large language models or AI as the same thing. You can effectively apply them to anything.

So I talk about Microsoft with the co-pilot products, which is essentially listening to your team's meeting and summarising it for you or turning a Word document into a PowerPoint presentation. But Adobe (NASDAQ: ADBE) has a thing called Firefly, in which you type words and it creates a picture for you for graphic design. You can think about it for pharmaceutical companies trying to work on drug discovery. You can think about it for banks trying to work out credit, Xero (ASX: XRO) for instance, helping you do your accounts.

So from our point of view, you should think about AI as a copilot or an assistant. It's not going to do your job for you, it's just going to make your job easier. And it's applicable across pretty much every industry in the world. So every company's going to have a go at using these models to create more productive tools. I suppose the big question we don't know is how successful they're going to be and how much we're all going to use them, but the early signs are really good.

However, it is important to distinguish between the high-end consumer and the low-end consumer. The high-end consumer is fairly unique in China. There's potentially even a dynamic where the high-end consumer is trying to get their money out of China - if that makes sense. So they're using high-end products to do that. And you see that in property markets as well in places like Melbourne and Sydney. So from our point of view, the jury's still out as to whether the high-end consumer in China is going to be affected by what's going on in the economy. But the economy's not doing well, and that's obviously not great for global growth and it is a concern that we have to monitor over the medium term.

The second thing to say is you don't actually need rates to fall from here. You just need them to stay the same, which is the way we think about the world. The whole Dot-Com bubble was inflated at 6% interest rates. It wasn't inflated at 0% interest rates. So this concept that you need 0% interest rates for growth equities to work is wrong. So we don't actually need rates to be cut, we just love them to stay the same. So from that point of view, that pressure has come off, the stocks we look at, and their earnings are starting to grow again and the stocks are starting to follow with earnings. So that's exciting.

The last thing I'd just say that I think a lot of people underestimate is central banks actually got rates to 5% or circa 5% without collapsing the financial system or without any major breakage. A lot of people thought this could never happen. A lot of people thought we would never get out of the post-GFC world and we have. And it's important to think about the long-term sustainability of the cycle from here because you now have central banks that have downside buffers that they could use. They've managed to get them in place without too much damage occurring. So I think that gets us much more excited about being able to take much longer-term views without the cycle getting in the way from here.

They said, 'We've still got inventory in the supply chain. Things are looking difficult.' And then by January, ChatGPT had come out, their phone's ringing off the hook and they're like, '2023 is going to be amazing.' And we're like, "Really?" And then they released the results in May and it was really amazing. And then they released their results in June and it was really amazing again. 

So just to put this in context, you've got a top 10 company on the planet that's covered by 28 analysts who've all got their earnings estimates for 2023 wrong by about 150%, and next year by more than 200%. So this has been a dramatic pickup in demand for NVIDIA's GPU products. We think it is just the beginning of the AI runway because the AI runway should last three to five years or at least. But is there an element of pull forward in that? Yeah, there could be. There could easily be an element of pull forward in that.

And I think that's what people are grappling with now. The stocks actually now cheaper than it was at the start of the year. So it's on 25 times next year's earnings. At the start of the year was on 50 times because of the huge earnings upgrade. And is that reflecting this pull forward or is this real? And I think the real question you've got to ask yourself on NVIDIA, which has come through on a number of the questions we've had here today, is how much are we actually going to use these products? 

If we end up using them a lot, and people would know this because they're already using them a little bit, they're probably using Bard or ChatGPT - how much are you going to use this product over the next three or five years? And if you think it's a lot, then it doesn't really matter from here for NVIDIA because NVIDIA is the winner. GPUs, they virtually have no competitors. They should be the big winners of this next computer shift, and we think you should just own it and not trade it for a long period of time from here.