Investors in Apple (AAPL -2.61%) have experienced anxious times recently. The company was feared to have fallen behind in the AI races due the lack of groundbreaking AI initiatives like those announced by its "Magnificent Seven" peers. Moreover, after a strong, five-year run, some may believe Apple is overvalued. Even Warren Buffett sold a significant 13% chunk of his stake in the iPhone giant during the first quarter.
However, Apple stock recently surged back to all-time highs after what appears to be a successful AI-oriented event this week. But do the new announcements make Apple a buy right now?
Apple's pragmatic approach to AI
Artificial intelligence (AI) has become a bit of a buzzword recently, with cloud-infrastructure giants spending incredible amounts of money on graphics processing units (GPUs) to build out their own capabilities.
However, Apple's consumer focus puts it in a different place. Remember, despite all of Apple's success, it has never actually invented any breakthrough technologies: Apple didn't invent the personal computer. It didn't invent the first portable MP3 music player. It didn't invent the smartphone. It didn't invent wireless headphones. It didn't invent the smartwatch.
What Apple does do is take existing inventions and perfect them with a unified hardware and software approach, integrating them seamlessly and naturally into how people actually behave, incorporating technology into their lives.
That appears to be the approach Apple is now taking with AI. And analysts seem to like it, as evidenced by the upgrades the stock saw the next day.
Apple Intelligence
Basically, Apple isn't so much rolling out a whole new AI service as much as infusing its products with homegrown AI models and those from third parties to make its existing iPhone and Mac franchises better. The new features fall under the name Apple Intelligence.
One key upgrade for Apple Intelligence will be Apple's personal assistant Siri, which will become much smarter and allow users to perform a lot of functions, through voice or text commands, that now take lots of searching and/or actions. For instance, you can now ask Siri to find a photo deep in the past based on a mere vocal description, or pull up the real-time flight information a friend may have sent you in a past email without having to scroll to find it. Siri will also be able to save addresses or messages someone sends you via iMessage via voice command, summarize or write notes across apps, and even create images based on voice prompts.
Basically, the gist of Apple Intelligence will be to make Siri an indispensable personal assistant that saves time and relieves users of everyday, time-consuming tasks.
And in Apple fashion, security is especially top-of-mind as a crucial concern in the age of AI. Apple disclosed most of these AI functions will be able to be done on the device. For those that require data being sent to the cloud, Apple has developed a new private cloud service it calls Private Cloud Compute. This will be an edge computing system that uses Apple's in-house designed silicon with new privacy capabilities, instead of expensive third-party GPUs in a public cloud.
And when one does want to use a third-party large language model (more on that later) that requires the more powerful models only found in public clouds, Apple will ask permission.
How will Apple monetize this?
Apple likely won't be selling AI services as add-ons but could benefit handsomely by spurring more iPhone sales via a faster upgrade cycle. The new AI features will only work on models iPhone 15 (last year's release) and above, which accounts for just 5% of the iPhone installed base, according to analyst estimates. While iPhone refresh cycles have been lengthening, these new AI features could be compelling enough to spur a wave of upgrades over the next few years.
Image source: Getty Images.
And Apple won't be spending hand-over-fist
Unlike other AI companies that have to build out massive large language models, Apple has taken a more capital-light approach, combined with partnerships with third parties.
For instance, consumers who want to use large language models with billions of parameters will likely need a massive and expensive model running on huge amounts of GPUs. But for this, Apple is outsourcing that to others. Apple Intelligence users will, for instance, get free access to OpenAI.
The OpenAI integration will happen later this year, while Apple also says it plans to integrate with Alphabet's (NASDAQ: GOOG) (NASDAQ: GOOGL) Gemini model in the future as well.
Will this cost Apple a lot of money or any money? It's hard to say. After all, there is a benefit for third parties to integrate with Apple and its massive, high-spending, global-customer base. For instance, Google pays Apple over $20 billion annually in exchange for Google being Apple's primary search engine on the iPhone. So, it's pretty unclear if Apple is actually paying OpenAI anything for the integration. Who knows? It could be vice versa, with OpenAI seeking to acquire users.
Apple is a safe stock but not super cheap
There is a pretty decent chance that large language models generating all the headlines today eventually get commoditized. Meanwhile, Apple has done a fantastic job of keeping focused and disciplined with AI, integrating it into its product offerings without spending huge amounts of money, while staying focused on use cases that truly matter to its customers.
At 32 times earnings, Apple is not a cheap stock. However, it's a cash-gushing company and an incredibly stable franchise, almost like a consumer staple stock. Those stocks tend to trade at high multiples but with stable growth resilient to economic cycles, inflation, and disruption.
Apple is the most consumer staple-like of the big tech stocks, which is why Warren Buffett likes it so much. Even the AI revolution doesn't seem like it will disrupt it.
