Apple's WWDC Reality: Scaled Down AI Ambitions As The iPhone Remains Dominant

Don't overreact to the AI headlines, Apple is still an iPhone company. Plus: A look at SpaceX's coming valuation.

When Tim Cook takes the stage at Apple’s developer conference in Cupertino today, he’ll be leading Apple through yet another moment of transition.

It will be Cook’s last WWDC as CEO as he makes way for incoming leader John Ternus. And it’ll be the second WWDC in three years to promise AI improvements to Siri and iOS as Apple transitions into its artificial intelligence era.

For Apple, the stakes to get AI right are high, and the headlines will likely overreact to what’s announced. With that in mind, here are two essential pieces of context to keep in mind as you take in the news: 1) Apple’s AI stumbles haven’t hurt its business at all so far, as the iPhone17’s sold like crazy and no competitive ‘AI native’ device has emerged, and 2) After a very ambitious Apple Intelligence keynote in 2024, the company’s ambitions have scaled down to where the technology can actually deliver today. Let’s dive into both:

Apple’s Business Is Surging Without AI

Despite the AI setbacks, Apple’s finances are in great shape. As of this morning, shares were up 55% year-over-year and up nearly 16% year to date. This is largely due to strong iPhone 17 sales: In the second quarter, iPhone revenue hit a new record and accounted for half of the company’s total $111.2 billion in revenue.

Even without a super-charged Siri, people can use top consumer AI apps like ChatGPT, Gemini, Claude, and Perplexity on their iPhones. And those apps’ underlying technology hasn’t (yet) been strong enough to power a dedicated AI device that competes with the iPhone (though OpenAI and a few others are trying).

Bottom line: Apple is still an iPhone company. And as long as the iPhone is selling, Apple is in good shape.

Apple’s Scaled Back Its AI Ambitions (And That’s Good)

When Apple announced Apple Intelligence at WWDC in 2024, it promised an almost-omniscient assistant that could understand your context and get things done for you. Infamously, it promised to be able to extract your mom’s flight details from your inbox, let you know when she’d be arriving at the airport by cross checking her confirmation email with available flight data, and tell you how long it would take to pick her up. Ultimately, Apple couldn’t deliver on the experience because its ambition outpaced the technology’s capabilities. The fallout led to legal liabilities. Just last month, Apple agreed to pay $250 million to settle claims over how it marketed AI features that weren’t yet available.

This year, Apple’s AI promises are much more in line with what AI models can actually do. There will be, for instance, a Search or Ask interface that will give you a way to chat with a selection of AI bots (Claude, Gemini, ChatGPT) simply by pulling down from the center screen.

Apple, effectively, will be leveraging its operating systems to deliver obvious AI improvements to its phone. And that might be enough, at least for now.

How Agentic AI Changes The Workplace — With Jacqui Canney and Kellie Romack (sponsor)

Kellie Romack is the Chief Digital Information Officer at ServiceNow. Jacqui Canney is the Chief People & AI Enablement Officer at ServiceNow. The two join the show for a facts-based conversation about AI automation in the workplace, recorded at ServiceNow's Knowledge conference in Las Vegas. Tune in to hear how each has automated work with AI, what the resulting shifts in their employees' work has looked like, and how managers should think about the technology. This is a evidenced-based, high-energy conversation with two leaders actually doing the work, a compelling look into what the future might bring for the rest of us.

The Intelligence Report

AI policy news:

• The CEOs of several top AI companies signed a new letter urging lawmakers to pass new requirements to help prevent AI models from being used to create bioweapons. The letter — signed by Demis Hassabis, Dario Amodei, Sam Altman, Alexandr Wang and Mustafa Suleyman — includes dozens of other signatures from top experts across tech, life sciences, biotech, national security and policy.

• President Donald Trump signed a new AI executive order asking tech companies to voluntarily give the government early access to frontier AI models.

• Senior U.S. government officials are exploring potential stakes in major AI companies including OpenAI, with Sam Altman reportedly the one who pitched the idea to the Trump administration last year.

• Separately, Bernie Sanders announced he plans to propose legislation that would create a sovereign wealth fund through a one-time 50% tax on tech companies’ stock, which he said will create more wealth for Americans and help “control the future of AI.” Recent reports said Altman has already met with Sanders to discuss possible public ownership, but without supporting the specific plan.

• In Congress, two House lawmakers released a bipartisan draft bill that would create a potential federal preemption of some state AI model regulations.

• A new law in New York State goes into effect this week (June 9) that will require disclosures on ads that feature AI-generated humans.

• The European Commission presented a new European Technological Sovereignty Package for the EU, including a proposed the Cloud and AI Development Act and Chips Act 2.0 as part of a broader plan to boost domestic production and decrease reliance on AI from American providers.

• The American Federation of Musicians of the United States and Canada sued Warner Music Group and Universal Music Group over their AI music licensing deals with companies such as Suno and Udio.

AI company and product news from last week:

• Online bot traffic surpassed human traffic for the first time in history, according to Cloudflare, which found bots now account for 57.1% of HTTP requests to select websites.

• OpenAI released more features for Codex for a broader group of users, including new role-specific plug-ins and ways to create websites for presentation.

• Google introduced Gemma 4 12B, a new multi-modal model designed to run on laptops while also offering stronger capabilities than smaller Gemma models.

• Microsoft released a number of updates at Microsoft Build, including a new cloud-to-chip platform called Project Solara to power an agent-first AI stack. It also demoed examples for how Solara could power devices and even employee badges.

• Meta released a new Business Agent for enterprises available across WhatsApp, Messenger, and Instagram. Other Meta-related AI news from last week: the company is reportedly delaying its Spark Muse model API for developers and reportedly exploring a major equity raise to fund AI infrastructure efforts.

• AWS announced a new $4 billion AI infrastructure deal with Pinterest through 2031, including Pinterest using AWS Trainium and Graviton to train and run models at scale.

Will investors buy SpaceX’s pricey IPO story?

When SpaceX likely goes public on Friday, investors might be paying a hefty premium for Elon Musk’s trifecta of moonshots.

SpaceX, which is targeting a $75 billion IPO at a $1.75 trillion valuation, filed its S-1 last month, giving a clearer glimpse of the company’s current financials. But at that price, investor returns might take even longer than Musk’s ever-moving timeline for reaching Mars. Last year, total SpaceX revenue was $18.7 billion, with $11.4 billion from Starlink, $4.1 billion from Launch and $3.2 billion from xAI.

SpaceX’s IPO price could be running ahead of its fundamentals, according to a new analysis by NYU finance professor Aswath Damodaran. He estimates the company’s equity value is closer to $1.3 trillion, or $100 per share, versus the reported valuation of $1.8 trillion, or $135 per share. Put another way, he said investors are effectively being asked to pay an extra $500 billion premium for SpaceX’s narrative that includes a long-term bet that launch dominance, Starlink growth, and xAI’s potential will pay off far beyond the company’s current financials.

The $28 trillion total addressable market outlined by SpaceX “borders on fantasy,” Damodaran wrote in a blog post while noting it’d be the largest TAM of any company in history. However, he added that the TAM estimation has been gamified by Silicon Valley and that AI agents could potentially be usable across most businesses and geographies.

The SpaceX story rests on potential revenues, target operating margins, capacity for reinvestment, Damodaran wrote. But if the core story is about growth potential and strong competitive advantages, the prospectus has made the narrative “bigger, but also more volatile” because the AI opportunity might be overstated and xAI’s competitive position still isn’t certain. He also noted potential issues with the governance structure that locks in Elon Musk’s control.

“If you add to this mix the antipathy that exists between Musk and Sam Altman, you have the potential for a UFC match between two monstrous egos, funded by tens of billions of dollars shareholder money,” Damodaran wrote.

The bull case is that SpaceX could be following Alphabet’s playbook in reverse, but with rockets. Just like Alphabet has used its wildly profitable ads business to fund “moonshots,” SpaceX seems to want public investors to fund actual moonshots before any future cash machine becomes concrete.

Right now, Starlink still brings the most cash, rockets bring cachet, and xAI brings the burn. But longer term, SpaceX’s prospectus shows enterprise AI as the biggest market story.