When reading posts, there are yellow stars against the names of the most respected Shrewds. The number of points in the star, starting at 3, represents the Shrewd'm-Star rating. This number is the average recommendation that the author received over the last 12 months.
- Manlobbi
Halls of Shrewd'm / US Policy❤
No. of Recommendations: 1
Manlobbi,
Would be interested to read your thoughts on Apple's recent WWDC announcements.
In particular, their recent rollout of (A)pple (I)ntelligence.
Will AI be a game changer for AAPL, or was this all form and no substance?
Does this change your view on future profitability for AAPL?
Thank You,
BHH
No. of Recommendations: 25
Will AI be a game changer for AAPL, or was this all form and no substance?
Does this change your view on future profitability for AAPL?
I'm quite sure will boost attention and a purchase cycle for iPhones for the Christmas cycle in late 2024. But as for long-term value, a few quarters is not important and doesn't effect intrinsic value much.
However the recent software improvements throughout the industry I am quite certain will increase value for consumers across the board. This point is subtle but important so I'll expand what I mean in this paragraph. I view all the AI buzz over the last few years as simply software becoming better. Unlike most, I don't view it categorically differently to all the incremental software advances over the last 20 years that also dramatically increased the value of using devices (the http protocol combined with hypertext, or the Word Processor and Spreadsheet for example, were each probably similarly important game changers as AI is now regard the incremental value). There was a lot of buzz about blockchain technology last year but that has mostly died away. In the past we just referred to AI as 'programming' along side all the other programming. I programmed robots at the age of 17 in which they learned from natural selection during deadly conflicts with each other, and gradually adjusted their behaviour. It produced one of the better selling Shareware games. But in those days ,as I wrote, we didn't call it AI and just called it "programming". When I later studied AI later, it was a deeper field of cognitive science where there was a real attempt to understand how we actually think - the science is still at an extremely early level today because attention has moved away from actual AI science research, to just engineering. It is now merely a statistical/engineering field, with no attempt to gain actual insight as to how we think, which is ludicrously more efficient than brute force engineering models we use now. Sometimes serious research and business profits are at odds with each other. The current 'engineering' AI direction is the right approach to find practical applications quickly, but regarding actual artificial intelligence the direction is a dead end as a science.
I think these software advancements will add value to how we use our phone, how we get results from Google searches, and elsewhere. If our phone is more valuable to us, we will be willing to commit more funds to maintaining the device. So I do think Apple's intrinsic value rises, but it isn't because of the recent Apple Intelligence announcement but because of software improvements throughout the industry increasing the value of all phones/tablets/laptops (both the devices themselves and the online services).
Keep in mind that if there is some new technology, in the past it often didn't increase profits because if one competitor started to use the technology (requiring capital expense), all the competitors followed - so they all had the capital expense, but then had no competitive advantage over each other anyway, and lowered their prices to similar levels before the technology was introduced. The result is lower profits because they were all just forced to endure the capital expense. So keep in mind that new technology applied across doesn't invariably increase profits. However, as I wrote, I think that AI really does increase the value of products, as it is not merely a manufacturing efficiency but something that the customer is able to benefit from.
To Apple Intelligence in particular, the part that I had not anticipated was their decision to enormously increase their R&D expense (because really the software is much more complicated) with focus on security using their own privacy chips at the server side. This shares their client-side privacy technology on the server, which auditors and observers have grown to trust. I think that was really smart, I had not anticipated it, and I expect it will give them some trust advantage over the competition, given how eerie it will feel to have the AI engine seeming to know everything about you.
I think Apple's intrinsic value should have some boost (yesterday's 7% seems reasonable), because the privacy using their own chip designs, and applying the AI to their ecosystem (your emails, contacts, messages, photos, music, etc) will be difficult for competitors to reproduce.
But the increase in value has to be set against their valuation. To reduce noise I'll use sales rather than earnings. Their price to sales (P/S) ratio is now 8.5x which is a record high versus its average of 6.3x the last 5 years (and an average of about 4.0x the 10 years before that). Much of the price increases that we observe for Apple are just the multiple gain. Sales per share increased 13.3x to 24.4x over the last 5 years, about 13% per year. Some of that has been from the buybacks. Over 3% per year of this 13% has been from the buybacks.
So they are growing on trend, before buybacks, at about 10% per year (at least 5% above inflation) and have a serious moat. Rather than paying a large dividend, they are dividing shareholder payouts between larger buybacks and a smaller dividend. I am not confident that they can continue something to grow at 5% above inflation before buybacks, because they are so consumer oriented and can only extract so much cash from customers each year. I suppose they have a longer runway for more laptop sales given that they have less market saturation there. I also believe they will continue to gain substantially higher earnings from continued growth in paid services from the premise that customers are not paying that much out for services today. But growing even 3% above inflation, before use of shareholder earnings, deserves a P/E multiple well above the the historical average of 15x for a typical firm. Apple's PE is 32x right now. Is that crazy high - I don't think so. Against its 20 year record it is very high, because it has historically traded cheap, but it is not objectively high given the moat. If it traded at 22x I would be really excited, and if it was trading at 40x I would think it is really overvalued. So I don't think the price is not on either of the extremes, despite it being at the high of its typical trading range. Boring summary, I know, but rare opportunities are, by definition, more... rare.
- Manlobbi
No. of Recommendations: 3
I'm quite sure will boost attention and a purchase cycle for iPhones for the Christmas cycle in late 2024. But as for long-term value, a few quarters is not important and doesn't effect intrinsic value much.
I disagree that this is just a "one time bump" - IMO this represents a new innovation cycle that will drive iPhone sales upgrades for the next few years at least. The technology is still VERY early; yet even with simple usage scenarios is generating a lot of excitement. I fully expect there will be a "killer AI App" that is launched in the next 18 months that will make all existing iPhones obsolete (in the minds of existing consumers).
tecmo
...
No. of Recommendations: 2
I wandered over here from the Berkshire board. I wholeheartedly agree with Manlobbi that real, general AI is a dead end as a science.
I’m reading a book titled “The myth of Artificial Intelligence- why machines can’t think like humans “ by Erik J Larson and he says the same thing. He talks of Deduction, Induction and Abduction. Abduction, loosely understood is what we think of as conjecture, common sense.
Manlobbi is correct, AI currently only statistical and nowhere close to attempting Abduction.
No. of Recommendations: 1
"Boring summary..."
Hardly. Thank you for for your insights.
No. of Recommendations: 3
This was a great post, Manlobbi. Thank you.
Agree that current AI engineering/programming. However, it is more powerful, esp. in the hands of Apple.
As tecmo wrote, which I had to read twice,
"I disagree that this is just a "one time bump" - IMO this represents a new innovation cycle that will drive iPhone sales upgrades for the next few years at least. The technology is still VERY early; yet even with simple usage scenarios is generating a lot of excitement. II fully expect there will be a "killer AI App" that is launched in the next 18 months that will make all existing iPhones obsolete (in the minds of existing consumers)."
Meaning that many existing Apple customers will need to upgrade to more powerful iphones (and computers) to access the inherent power of the Apple eco system enhanced by this new AI alliance. Apple being the preeminent operating system/device eco system in existence. (Sometimes I need to reword things so I can understand them better.)
See the link from Cardude on the Berkshire Board, today, copied and pasted below,
https://stratechery.com/2024/apple-intelligence-is...An exhaustive article on the opportunities in the Apple/ChatGPT alliance as well as its possible disruptions from Microsoft and/or Google.
Good information there.
Also, very much appreciate the Apple valuation estimate.
--Elizabeth
No. of Recommendations: 1
Meaning that many existing Apple customers will need to upgrade to more powerful iphones (and computers) to access the inherent power of the Apple eco system enhanced by this new AI alliance. Apple being the preeminent operating system/device eco system in existence. (Sometimes I need to reword things so I can understand them better.)
Yup, we already know that existing iPhone customers upgrade on a regular cycle, this is likely to accelerate and protect the next few upgrade cycles. I wouldn't be surprised to see a margin expansion as well.
FYI: There already are a few compelling AI Apps - of course ChatGPT is the most obvious one, but there are some that will generate music, photos, etc... These are all "toy" apps right now, but new ones are coming that will be much more interesting.
tecmo
...
No. of Recommendations: 0
What was interesting to me was the interactivity of the various AI apps with the Apple OS, so one could integrate them into the basic operating system.
No. of Recommendations: 2
What I mean is, as an example, in the the old days, when Apple was a player in film editing, using Adobe products and then its own, various software was used for compiling, special effects, music editing and animation, which could interact with the basic editing software. They played well together. So, it looks like that is what is on the table for Apple OS and the apps that can be developed for Apple OS and AI. The wider reach the os has, the more complex and robust the web of applications for everyday AI can be. Thankfully, AI, in these instances, are, in fact, software, but more and more seamless. The matter of safety and personal control will be crucial, which Apple is best at, right now.
--my interpretation, from the article posted earlier, in Stratechery.
No. of Recommendations: 2
What was interesting to me was the interactivity of the various AI apps with the Apple OS, so one could integrate them into the basic operating system.
I think AI will generally be part of the platform - sort of like how GPS or the Camera is. Then new Apps will hook into that to create unique offerings.
tecmo
...
