Volatility is back.
Q3 brought many wild swings back into the market over earnings, election rumblings, and geopolitical conflict. The much-loved semiconductor index finally saw a pullback after its relentless run looked nearly unstoppable. Investors sharpened their pencils and finally asked the grand question: Does this all make sense?
As Nvidia re-writes the stock market history books with the largest beats and raises in recent memory, the torrid pace of growth seems now much more predictable. Sell-side analysts are no longer off on their models by orders of magnitude, and the delay of the Blackwell series finally put a question mark on a company that has amassed rabid fans.
However, Nvidia’s revenue predictability does not make it less impressive. They are putting up massive revenue numbers at extremely impressive margins, which shows just how much of a stranglehold Jensen has over the entire semiconductor supply chain.
I want to discuss two ideas that are trending in the technology sphere and consider what this means for the entire space. There are so many technical things happening very quickly with AI, and you have to zoom out, take a wider approach, and try to figure out how it is structurally changing business models and competitive dynamics.
The first advancement in the AI world was the model OpenAI most recently released called o1 (nicknamed strawberry). They have renamed a new run of this model because they think of it as a day-one technology. Essentially, this model applies reasoning to specific questions; it actually thinks through its own answer. This is done through chain-of-thought prompting. Previously, we had a narrowly super-intelligent information retriever, but now we have reasoning.
This has been a breakthrough when it comes to the STEM fields. In a qualifying exam for the International Mathematics Olympiad, o1 scored 83% accuracy, compared with GPT-4o's 13%. What we're getting here is a model with incredibly enhanced capabilities across more technical fields.
You may be familiar with it, but there was a famous meme where if you asked any of these LLM chatbots how many letter Rs are in the word strawberry, it would have a total meltdown. This model actually gets that question right, hence the nickname. But what is important is the process by which it arrives at the answer. It does this through thinking and reasoning for much longer, which unlocks applications for entirely new fields.
That’s the what happened.
As for the why this matters... this whole thinking process will do two things. The first is that it will unlock entirely new applications. For example, think about a very complex problem like Tesla and the snags that they're having right now with their 4680 battery production. You can run physics problems in these advanced models that can solve manufacturing problems. Thinking through and reasoning with these types of models will lead to scientific breakthroughs.
We saw how chunky and clumsy GPT-3 was compared with GPT-4o, and the same thing is going to happen with these reasoning models. This has applications across every industry that can benefit from just how much smarter the reasoning is at technical problems. For those that adopt these techniques, the result will be declining R&D costs as well as revenue growth acceleration as entirely new products are built.
The second implication as to WHY this matters is that this thinking process requires a ton more inferencing. Remember, in AI, there are two parts to these models. The first is training, AKA building the model. The second is inferencing, AKA calling the model. The reasoning version of the model is going to require a lot more inferencing, which rapidly increases the demand for more infrastructure investments.
More inference means higher demand for NVDA chips, data centre cooling, data centre REITs, and nuclear energy to power these systems.
More guts, more glory
What we have here are power laws coming into effect that are really accelerating the capabilities of these models, and they're only going to improve over the medium term.
The thesis here is that we are not done yet with the infrastructure layer – we are still only getting started.
Another big idea circulating in tech circles is that these new LLM creators have the potential to be the next version of "Big Tech" over the next decade. Foundation model providers such as OpenAI, Anthropic, and Perplexity have the potential to enable new businesses based on their own increased API adoption.
Some VCs compared these model providers to the early days of Cloud providers. As Cloud adoption led to new businesses in Cloud Security, Cloud Data Management, and Cloud Observability, VCs believe AI LLM adoption could unlock new and yet-to-be-seen AI-only business opportunities.
As to WHY this matters – we have to think about the dominance of some of these large players in the industry. Even though Google invented the transformer technology that is changing the game, their core business of search is under attack. These days, I'd rather turn to an answer engine rather than a search engine, and if I need a product, I search natively on Amazon instead of on Google. Amazon has quietly built a $50B/year business in the ad space.
When it comes to who is powering this revolution, the cloud players are providing the pipes. AWS and Azure are leading this charge in a big way and focusing on their developer communities, with a lot of this closed source. Additionally, there is the darkhorse with Meta, which is doing this in an open-sourced fashion and monetizing later.
These players will swap leadership positions back and forth as products are built out on their systems, but the idea that the next new companies are all being built with the major LLMs as the guts is a very interesting concept that will have broader implications across the ecosystem that could disrupt the order of big tech.
New guts, more glory?
While I do believe that the incumbents will strongly benefit from their cloud businesses in the early build-out stages, we also have to look on the horizon for what the next "big tech" theme is going to be. I believe it will be applications built natively on top of LLM providers. This will disrupt many traditional SaaS models and is coming over the medium term.
We also have to focus on the stages of deployment. The big trade that everyone has been on is this infrastructure build-out, and over the course of Q3, what we had was a bit of a reversion in that trade as some of the semis sold back off, and so did the supporting players like Vertiv, Dell, HPE, etc… but we're starting to see some of these stocks come back now. I don't think that this buildout phase is over and we could have another run at it with the infrastructure names here. So staying long the supporting system and energy plays here is a great bet. In addition, look for names that can immediately monetize AI capabilities to outperform.
Wrapping Up
I believe now is the time to start looking at this theme, broadening out, and seeing which players are using AI in which industries, such as manufacturing, drug discovery, supply chain management, and other deep-knowledge fields.
These reasoning models will be copied and improved on very quickly, and they will start making discoveries that human intelligence couldn't achieve.
Strong Convictions. Loosely Held
Nick Mersch, CFA
Get the latest tech insights to your inbox every month.
The content of this document is for informational purposes only and is not being provided in the context of an offering of any securities described herein, nor is it a recommendation or solicitation to buy, hold or sell any security. The information is not investment advice, nor is it tailored to the needs or circumstances of any investor. Information contained in this document is not, and under no circumstances is it to be construed as, an offering memorandum, prospectus, advertisement or public offering of securities. No securities commission or similar regulatory authority has reviewed this document, and any representation to the contrary is an offence. Information contained in this document is believed to be accurate and reliable; however, we cannot guarantee that it is complete or current at all times. The information provided is subject to change without notice.
Commissions, trailing commissions, management fees and expenses all may be associated with investment funds. Please read the prospectus before investing. If the securities are purchased or sold on a stock exchange, you may pay more or receive less than the current net asset value. Investment funds are not guaranteed, their values change frequently, and past performance may not be repeated. Certain statements in this document are forward-looking.
Forward-looking statements (“FLS”) are statements that are predictive in nature, depend on or refer to future events or conditions, or that include words such as “may,” “will,” “should,” “could,” “expect,” “anticipate,” intend,” “plan,” “believe,” “estimate” or other similar expressions. Statements that look forward in time or include anything other than historical information are subject to risks and uncertainties, and actual results, actions or events could differ materially from those set forth in the FLS. FLS are not guarantees of future performance and are, by their nature, based on numerous assumptions. Although the FLS contained in this document are based upon what Purpose Investments and the portfolio manager believe to be reasonable assumptions, Purpose Investments and the portfolio manager cannot assure that actual results will be consistent with these FLS. The reader is cautioned to consider the FLS carefully and not to place undue reliance on the FLS. Unless required by applicable law, it is not undertaken, and specifically disclaimed, that there is any intention or obligation to update or revise FLS, whether as a result of new information, future events or otherwise.
