Apple may soon need to make up its mind about spending on AI.
The smartphone giant has never really been all-in on the technology. It resisted developing its own AI models, relying instead on OpenAI to fuel its AI strategy under a 2024 deal.
Earlier this year, it abandoned OpenAI in favor of Google's Gemini, which is set to power a revamped AI-infused version of its Siri digital assistant.
There have been advantages to this approach. Notably, it is capital-light. Instead of absorbing a huge increase in capital spending to fund model development with an uncertain outcome, Apple is passing that financial pain along to suppliers. It may pay more, compressing its margins, but it risks less.
So far, that has looked shrewd. Despite rising computer-memory costs and a tight supply chain that have been constraining phone sales, the company is prospering. Its shares are down about 10% since its earnings last week exposed those constraints. But its stock is still up more than 13% this year, and it is the second-most-valuable listed company in the world.
But as the switch from OpenAI to Gemini illustrates, there are also disadvantages to its less-than-absolute commitment to AI. Choosing the wrong AI model supplier has consequences.
And Apple doesn't own its own AI destiny. This is philosophically unusual for a company that prides itself on its vertical integration and tight control over its devices and software.
Siri AI, set to roll out to the public this fall, may finally force Apple to commit more strongly to AI, even if it doesn't dip toes into model development.
While Apple's phones and computers can handle many AI tasks themselves, Siri AI's capabilities require more computing horsepower than they can provide. That means Apple will need to add to its AI computing capacity if the AI assistant takes off.
Asked about the capital intensity of running Siri AI, Apple chief Tim Cook told analysts last week that it was using a mix of its own data centers and leased cloud-computing services to underpin the AI strategy. "We'll see what Siri AI does" to the company's costs, he said.
Apple, Cook said, was absorbing a lot of AI costs in its operating expenditures and other entries in its income statement, implying the company might not need to grow capital expenditures much. Subscriptions to more-expensive versions of its iCloud service for heavy AI users could help offset the costs.
That would be a coup at a time when many investors are wary of companies growing AI spending. But it isn't clear whether threading that needle is possible if Siri AI is a runaway hit.
Ultimately, success in AI requires spending -- even for Apple.
This is an edition of the WSJ AI & Business newsletter, a weekly digest to help you make sense of AI's impact on business with news, insights and data from our global team of technology journalists. If you're not subscribed, sign up here.
How OpenAI Lost Its Crown
OpenAI was the AI boom's first darling. Not so much any more. Sam Altman's company is bleeding cash and falling behind its rival Anthropic as the companies race toward IPOs. Anthropic vaulted ahead of OpenAI by seizing on a lucrative niche -- developing world-leading AI coding tools -- instead of looking for a foothold in mass-market AI where it would challenge Google and other tech heavyweights. Now, Anthropic is pushing ahead with plans for a fall IPO, while OpenAI may wait until next year.
The Number
The number of active users OpenAI surpassed after cutting prices for several models as it tries to gain ground on Anthropic.
What the Humans Are Saying
AI in Charts
Amazon.com has dominated the cloud-computing race for two decades.
In the AI era, that may be changing. Atop the cloud-computing industry are two other companies that are increasingly its equal: Microsoft and Google.
Amazon is holding its position at the top for now, with revenue growth in its cloud business that is accelerating. It rose to 37% growth in the second quarter, easily beating Wall Street projections.
But Microsoft and Google are also growing, and faster. Google's cloud business grew more than 80% in its most recent quarter. Microsoft's grew at over 40%, aided by big cloud-computing deals with AI labs.
Amazon chief Andy Jassy has been dismissive about the competition in the past. In February, he said it was "very different" to have a lower growth rate on his much larger business "than to have a higher percentage growth on a meaningfully smaller base, which is the case with our competitors."
Yet if that trend continues, it isn't impossible for Microsoft and Google to get close to or even surpass Amazon's cloud-computing revenue within a few years. That would be a big shake-up for an industry that hasn't historically seen a rapid market-share shift.
AI in the Wild
For better or (probably) worse, AI is making its way into books, rocking the publishing industry in the process. A debut author shopping a manuscript for a book called, "Call Me, I'll Hide the Body," recently sparked a bidding war that ended with a $2 million, two-book deal. But the author's agent subsequently pulled the book amid concerns about AI use, making it the latest in a string of such publishing-industry scandals.
Other Highlights From the Week in AI
-- Visa is paying $2.4 billion for a company that uses AI to detect fraud. -- Tesla is considering selling its China unit to prepare for a possible merger with SpaceX, which holds Elon Musk's AI businesses. -- Anthropic's AI models hacked three companies during tests, following a similar incident involving OpenAI and HuggingFace.
About Us
WSJ AI & Business is a weekly look at AI's transformation of the business world. This newsletter was curated and edited by Asa Fitch. Reach him at asa.fitch@wsj.com (if you're reading this in your inbox, you can just hit reply). Got a tip for us? Here's how to submit.
Comments