Apple will need to make the biggest acquisition in its history to put an end to its artificial intelligence debacle. Also: Meta rolls out new Oakley smart glasses; Honor prepares a new ultrathin foldable phone; and Apple’s annual back-to-school deal gets underway.

Last week in Power On: Apple’s upgrades to CarPlay, the iPad and Vision Pro outshine its new Liquid Glass interface.

The Starters

Apple AI head John Giannandrea and acquisitions chief Adrian Perica (center) having a conversation before WWDC. Image Credit: Ryan Isaacs

Among the giants of Silicon Valley, Apple Inc. has long had an unusual acquisition strategy. The company’s biggest-ever purchase was Beats in 2014 — for $3 billion — and Apple has only made three transactions totaling $1 billion or more in its entire history.

Rather than acquiring large, transformative businesses, Apple tends to target small companies with focused teams — typically to accelerate an internal project or jump-start a new initiative.

Compare that with Facebook’s $19 billion purchase of WhatsApp, or Google’s $12.5 billion deal for Motorola Mobility. Just a couple years ago, Microsoft Corp. plunked down $69 billion for Activision Blizzard.

When Apple has attempted billion-dollar deals, they haven’t gone smoothly:

• The Beats transaction led to a culture clash that took years to resolve. Still, the acquisition ultimately paid off for the company by laying the groundwork for Apple Music and bolstering hardware sales.

• The $1 billion purchase of Intel Corp.’s modem business brought in patents and engineers to fuel Apple’s custom chip ambitions. But much of Intel’s talent didn’t mesh well at Apple, and large portions of the acquired technology were abandoned. Apple’s first in-house modem chip, dubbed C1, arrived years behind schedule.

• In 2016, Apple invested $1 billion in DiDi, then the “Uber of China,” hoping to strengthen ties in the country and potentially support its secretive car project. The bet imploded after Chinese regulators cracked down on DiDi, leaving Apple’s investment virtually worthless.

These experiences have made Apple deeply skeptical of large-scale M&A. Moreover, Apple’s homegrown approach often works out fine:

• Wall Street long urged Chief Executive Officer Tim Cook to buy Netflix Inc. or a major movie studio. Instead, Apple built the TV+ streaming platform from the ground up. That service is now on a path to long-term success.

• Apple also resisted calls to acquire Tesla Inc. or Rivian Automotive Inc. While not scooping up Tesla in its early days was arguably a missed opportunity, avoiding a more recent acquisition was a wise move — especially since Apple ultimately scrapped its car ambitions altogether.

• Before Apple Music launched, some advocated for buying Spotify Technology SA or Pandora Media. Though a Spotify acquisition might have jump-started Apple’s audio streaming efforts, the company was able to build up the service on its own. Pandora’s radio-style model, meanwhile, is largely obsolete.

That said, Apple is facing a very different set of circumstances these days and may have no choice but to do a deal. The stakes of the artificial intelligence race are just too high.

Developing the next generation of products — from glasses to robotics to new wearable devices — will hinge on AI. The technology will be as foundational to those products as the multitouch interface was to the iPhone (and that, notably, came via Apple’s 2005 acquisition of FingerWorks).

Apple’s in-house artificial intelligence work has floundered. Sure, the company has its talking points: Apple Intelligence is smoothly integrated, and rivals don’t protect privacy the way it does. Who really needs a separate chatbot anyway? But the truth is clear. Apple has missed the AI moment — and it should accept that instead of making excuses.

If Apple had built something as groundbreaking as OpenAI’s ChatGPT, it would be celebrating it from every rooftop. Instead, the company is downplaying the competition and branding its offering as “AI for the rest of us” — because that’s really all it can say. Apple is doing its best with what it has and will continue to shrug off any concerns until it can offer something better.

Aravind Srinivas, chief executive officer of Perplexity AI, during the Bloomberg Tech conference in San Francisco. Photographer: David Paul Morris/Bloomberg

The good news is, Apple has a cash hoard of more than $130 billion and a way out: It can buy its way into the future. The right acquisition could bring in the talent and technology Apple has clearly lacked — as evidenced by the struggles of its Siri voice assistant and its underwhelming generative AI work. This time, transformational M&A is exactly what the company needs.

But first, let’s rule out some possibilities. OpenAI and Anthropic PBC, two of the leading AI developers, are not likely targets. OpenAI has a $300 billion valuation, as well as a relationship with Microsoft. Its convoluted corporate structure also makes it an impractical acquisition. Anthropic isn’t much more attainable. It’s valued at around $60 billion and has deep backing from Amazon.com Inc. and Google. Both deals would invite serious regulatory scrutiny.

The most likely acquisition target in the multibillion-dollar range is Perplexity AI. While the company lacks its own cutting-edge foundation models — like those of OpenAI or Anthropic — it has several advantages that make it a logical fit for Apple:

1. A proven, consumer-ready product. Perplexity’s search tool — with its text interface, voice controls and sleek design — already feels at home on Apple’s iOS. It could easily become the default AI-powered search engine for the Safari browser, Spotlight feature and even Siri.

2. It fills a clear need. Apple is focused on improving the AI foundation models that power its platform. What it really lacks is a strong search layer and a conversational interface for everyday tasks — exactly where Perplexity excels.

3. A decently sized team. Perplexity has roughly 250 employees, meaning it’s not too big for Apple to integrate. The AI talent is up there with other industry players and could be useful across the company’s Siri and AI engineering teams.

4. Reasonable valuation. At around $14 billion, Perplexity wouldn’t break the bank. That’s a fraction of what Apple would have to spend for other major AI players, making it a relatively low-risk, high-upside move.

5. Timing. Apple’s long-running search deal with Google is under threat from a US antitrust suit. The iPhone maker has to prepare for a post-Google search world — and acquiring Perplexity would allow it to finally launch its own Apple-branded search engine.

Such a deal might help in other ways. Perplexity has quickly become one of the most recognizable brands in AI, and bringing it in house could give Apple more credibility and recruiting power in the AI community — a space where it has historically struggled to attract top-tier talent.

And Eddy Cue, known as Apple’s key recruiter and dealmaker, has already publicly declared his interest in the startup. “We’ve been pretty impressed with what Perplexity has done, so we’ve started some discussions with them about what they’re doing,” Cue said in his testimony at the Google antitrust trial earlier this year. He went on to say the service sometimes can provide better results than a classic search engine like Google.

Apple executives, including Cue, deals chief Adrian Perica and AI managers, have internally debated whether it makes sense to buy Perplexity, I reported on Friday. The companies aren’t in active or formal acquisition talks, though. For now, Apple is focused on evaluating Perplexity’s technology and has met with its leadership team.

Apple is still weighing how much of its AI arsenal it wants to build, buy or license — and it’s waiting on the courts to determine the fate of its Google search arrangement before picking up talks.

Separately, Apple met earlier this year with Mira Murati — the former chief technology officer of OpenAI — to discuss a potential deal for her new AI startup, Thinking Machines Lab. The talks never progressed to an advanced stage.

There are other small AI companies without sky-high valuations that Apple could consider, including Cohere, Sierra AI, Databricks and Mistral.

Unlike the others, Mistral would give Apple a serious leg up in building cutting-edge foundation models. Its high-performance models are widely respected for their efficiency and speed — ideal for running on devices or in the cloud.

Apple’s current models are relatively underpowered. Its on-device model has 3 billion parameters — a measure of complexity and learning ability — while its cloud version has 33 billion. That’s far behind what competitors like OpenAI and Google are deploying.

Buying a multibillion-dollar AI company would represent a radical shift away from Apple’s conservative M&A strategy. But sticking to business as usual risks leaving the company even further behind in the most important tech race since the smartphone.

The Bench

The Oakley Meta HSTN glasses. Source: Meta

Meta enters the sports-centric smart glasses market with new Oakley model. The market leader in smart glasses — thanks to its hit Ray-Ban models — is adding a version with the Oakley brand. The new product, based on the Oakley HSTN design, will be geared toward athletics and bike riding. It goes on sale this summer for $399, with a limited-edition, gold-accented version arriving earlier for $499 (this one is available for preorder on July 11).

These new Meta Oakleys look fantastic to me, and I absolutely cannot wait to try them out (look for a full Bloomberg review as soon as possible). The features are familiar if you’ve tried the Ray-Ban version. They have cameras to take pictures, shoot video and analyze the environment with AI, plus the glasses can take phone calls and play music.

But the Oakley version has a few extra bells and whistles:

• Battery life has been doubled from four hours on the Ray-Bans to eight hours. They also charge up to 50% in 20 minutes, and the charging case allows for up to 48 hours of use.

• The camera is higher-resolution, allowing for Ultra HD (or 3K) video capture.

• It has a water-resistance rating of IPX4, which means it should be able to survive water splashes or rain.

With still more than a year to go before Apple enters the smart glasses market, Meta has a hit on its hands and is already offering a second brand. The company is also preparing to launch a higher-end pair with a display later this year. Apple will have a steep hill to climb next fall, especially because its AI is inferior to Meta’s — despite the social media giant’s own struggles in this area.

An Honor Magic V3 foldable smartphone. Photographer: Angel Garcia/Bloomberg

Honor’s next foldable phone will be just 8.8 millimeters thick. Honor, a top Chinese phone maker, is skipping the Magic V4 — the would-be name of its next foldable phone — and will instead launch one next month called the Magic V5. The new model will come in at 8.8 mm when closed, Honor tells Power On. That means it will very likely be the world’s thinnest foldable phone when it debuts. The V2 came in at 9.9 mm, while the current V3 measures 9.2 mm. It is believed that the latest Honor phone will be thinner than Samsung Electronics Co.’s upcoming Z Fold 7, which will also be introduced in July. Apple, for its part, is poised to launch a foldable at the end of 2026.

US Retail Sales Unexpectedly Rise In Solid Start To Holidays

Shoppers at an Apple store in Walnut Creek, California. Photographer: David Paul Morris/Bloomberg

Apple’s back-to-school promotion gets underway. As expected, Apple’s annual educational sale kicked off this past week in the US. The idea is to drive demand for Macs and iPads ahead of the back-to-school shopping season that usually kicks off in the middle of the summer. This time of year is typically a slow sales period for Apple — coming ahead of its annual iPhone launch — and a good time to boost sales of computers and tablets.

With the latest deal, there’s a bit of a twist: Instead of offering gift cards for services and app downloads, Apple has pivoted to giving away some accessories. An iPad, for instance, gets you a free Apple Pencil Pro. A MacBook comes with noise-canceling AirPods 4 or a wireless keyboard. The iMac also entitles you to free AirPods. If you want other accessories — like a Magic Keyboard for the iPad — you can get that too, but at a discount rather than for free.

This is a fascinating strategy, and there’s more than meets the eye:

• By getting Apple accessories, customers are now locked into a product ecosystem that — the company hopes — they’ll fall in love with and never leave.

• The Apple Pencil Pro requires an iPad Pro, iPad Air or iPad mini, which could encourage someone to buy something higher-end rather than a regular iPad.

• The accessories that carry a discount (but aren’t free) may prod customers into spending more money with Apple than they would if this program didn’t exist.

• The MacBook, not iMac, is the device that gives you a free trackpad, mouse or keyboard. That’s curious since many laptop users don’t need those accessories. But it could set them up to eventually buy another Mac — an iMac, Mac Studio or Mac mini, for instance — in the future.

• Customers who try AirPods often have a hard time giving them up. That means they’ll buy new ones if the accessory gets lost or the battery wears out. Luring someone into the AirPods universe with a free initial pair is probably going to guarantee a series of purchases in the future.

So while a free or discounted accessory may just seem like a good deal for the customer, it’s actually a carefully crafted incentive aimed at boosting sales in the long run.