Apple AI - As Good As it Gets

Investment Note #15 - 21st June 2024

Apple AI - As Good As it Gets

Synopsis

• In the 32 trading days to the 5th June 2024, Nvidia added more than $1 trillion in market capitalisation, with this 6-week gain being greater than the total market capitalisation of Berkshire Hathaway which Warren Buffett has spent six decades building (source: Bloomberg).

• Aswath Damodaran is a well-known Professor of Finance at the Stern School of Business and is widely respected for his work on company valuations. Recently on CNBC, he commented on Nvidia’s $3+ trillion valuation stating that if you were designing a perfect momentum company from scratch, then Nvidia would be it; a great story, a CEO who sticks to the story and a receptive market environment.

• Concerns about a possible AI-fuelled bubble tend to focus on Nvidia’s meteoric share price rise. We ask if it would be more instructive to consider that question in the context of a company we are all more familiar with – Apple.

Corporate Partnerships

• The following is an extract (with one word blanked out) from a speech by the CEO of a leading technology firm: “This is a pretty interesting opportunity and I think we will see a lot of different people actually going after this trying to figure out how to make [BLANK] turn into something that is actually useable and mission critical and can actually deliver ………mission-critical kinds of services…..building the design centre for all the products, technology, the Application Programming Interfaces (APIs) and partnerships, assuming that will happen……….”

• In the audience at Apple’s recent Developers Conference, the annual Apple jamboree which is the forum for unveiling what the firm is working on, was OpenAI’s CEO Sam Altman. If we said he joined Tim Cook on stage and made the above statement about the two companies working together, with the BLANK being generative AI, you might well believe us.

• You would be wrong to do so. The BLANK was the World Wide Web, and this was a speech in 2000 from Sun Microsystems CEO Scott McNealy praising Enron’s Internet strategy. McNealy attended the meeting after Enron agreed to buy 18,000 servers from Sun Microsystems. McNealy’s appearance was designed to help convince the analysts present that Enron had partnered with a tech sector heavyweight and the announcement of the $100m order from Enron had the desired effect – Enron’s share price was +24% on the day (source: University of Missouri). 

• Now, we are not suggesting that Apple partnering with OpenAI is comparable to Enron partnering with Sun Microsystems, but we are suggesting that there may be some similarities between that period and now, that at the very least, may give pause for thought. It is surely interesting that given the generative AI momentum, Apple felt it necessary to announce that it had partnered with OpenAI to integrate ChatGPT into its devices. As Gene Munster of Deepwater Asset Management was quoted in the Financial Times, “they’ve never gone to a third party to license a core technology”.  Compare this to Apple’s Developers Conference in 2023, in all the speeches made by senior executives generative AI wasn’t mentioned once. Between the two conferences, Nvidia added c. $2 trillion in market capitalisation and Apple added c. $100 billion. On 13th June 2024, two days after Apple’s OpenAI announcement, Apple added 10% or $305 billion in market capitalisation (source: Bloomberg).

Tech’s ‘Meh’ Moment

• Unlike many iPhone users, the Key Capital investment team are big fans of Siri and can often be overheard strolling down Merrion Row saying “Hey Siri, forward 60 seconds” to skip annoying podcast advertisements. But even coming from that position of Siri fandom, the Apple announcement seems somewhat underwhelming. A significantly improved Siri may be useful but is it a transformative advancement?

• Maybe we are being unfair to Apple and maybe it is part of a broader challenge facing the tech sector. Every major tech company is ‘integrating AI’ into their products and services, yet underneath the hood, the ‘AI’ they're integrating doesn't seem to do anything particularly new or solve any particular need. So far, it looks like a solution in search of a problem.

• But we shouldn’t forget that Apple’s strength has been implementation rather than innovation. It didn’t invent the mp3 player, tablet or smartwatch but it perfected them. Could the use of a smaller large language model inside Apple be a better fit for the needs of consumers?

• Indeed, will it transpire that a generative AI assistant that fully understands an individual’s world intimately and can harness that knowledge via a large language model could actually be the ‘killer’ generative AI application? Apple Intelligence’s integration of ChatGPT may become the consumer’s greatest use of AI but while this may be great for Apple, what will it mean for the AI capital expenditure and AI-related share price appreciations of other firms?

• Could it be that Apple’s AI announcement and its future AI success, if it delivers on it, will be the starting bell for figuring out who the AI winners and losers are. If the winner is going to be arguably the most consumer-focused business in the world, could the loser(s) come from the ranks of those businesses further removed from the consumer?

The Risk of Building Out Capacity

• Bernstein Research have drawn the comparison between server sales during the dot-com era as a precedent for the growth acceleration that Nvidia is currently experiencing. The Sun Microsystems experience also highlights that upstream hardware vendors often experience more demand volatility than businesses focused on end-user demand. Apple’s announcement aside, as major user-facing AI applications are yet to appear, the revenue-generating capabilities of these applications are a significant unknown. Sun Microsystems market capitalisation peaked at $200 billion during the dot-com boom, only to have the company acquired by Oracle Corp in 2009 for under $6 billion (source: MarketWatch).

• Another historic buildout of capacity that didn’t go to plan was the Lucent, Nortel and the broader fibre buildout in the 1990s. As Bernstein highlights by 2001, companies had laid c. $90 billion worth of fibre, and only 2.6% of that capacity was in use. Indeed, even to this day, there remain tens of thousands of miles of fibre that is still ‘dark’ or ‘unused’.

• According to Giuseppe Paleologo, an intern at Enron in the summer of 2000, the Sun Microsystems servers were languishing in a storage facility in Houston that summer.  According to eWeek, the servers were offered for 6 cents on the dollar in 2002 as part of Enron’s fire sale of assets.

• In a recent post on X, formerly Twitter, Elon Musk confirmed he had reallocated chips destined for Telsa to two of his other companies, X and xAI. In a comment that echoes with the past, he stated otherwise “they would have just sat in a warehouse”.  

Summary

• The AI revolution is still in its infancy, and none of the Big Tech companies have yet managed to graft generative AI onto a killer hardware or software product. Maybe the first question to ask the new AI super-charged Siri is: “Hey Siri, are we in the middle of an AI fuelled bubble?"