Will Apple (Finally) Get AI Right At WWDC?, Anthropic’s Worry, Microsoft vs. OpenAI

Summary of Will Apple (Finally) Get AI Right At WWDC?, Anthropic’s Worry, Microsoft vs. OpenAI

by Alex Kantrowitz

57m•June 5, 2026
Listen to Episode

Overview of Big Technology Podcast — Will Apple (Finally) Get AI Right At WWDC?, Anthropic’s Worry, Microsoft vs. OpenAI

This episode covers three major AI stories: Apple’s upcoming WWDC and its renewed Siri strategy, Anthropic’s warning that AI may soon improve itself too quickly for humans to control, and Microsoft’s increasingly direct competition with OpenAI. The discussion is skeptical, but nuanced: Apple may finally be aligning its AI plans with what the technology can actually do; Anthropic’s safety messaging may be partly real and partly marketing; and Microsoft appears to be rethinking how dependent it wants to remain on OpenAI.

Apple WWDC Preview: Siri, AI, and the iPhone as the Interface Layer

The first segment focuses on Mark Gurman’s reporting about Apple’s WWDC announcements, especially a revamped Siri.

What Apple is reportedly building

  • A new Siri, internally known as Project Campo
  • A more capable assistant that can:
    • launch apps
    • send texts
    • check weather
    • add calendar events
    • search notes
    • trigger shortcuts
    • help search the web with AI
  • A new “search or ask” interface accessed by swiping down from the top of the iPhone
  • A possible dedicated Siri app for ongoing conversations and chat-style use
  • Support for multiple model options via a model switcher
  • Under the hood, Apple is reportedly relying on a Google Gemini-based model, with some of the system hosted on Google servers

Why the hosts think this matters

  • Apple may finally be treating AI as an interface problem, not just a model problem
  • The iPhone gives Apple an enormous advantage because it already owns the operating system and the primary user touchpoint
  • Even if OpenAI, Meta, or Google build better agents, Apple can still embed AI directly into the device experience
  • The hosts see this as Apple using its installed base and inertia to stay central in the AI race

Concerns and skepticism

  • The features being described are fairly basic and feel like table stakes in 2026
  • Apple has not yet proven it can execute reliably on AI
  • The Siri rollout may be labeled beta/preview and could include a waitlist, which raises concerns given Apple Intelligence’s rocky launch
  • There are privacy questions around Apple leaning on Google infrastructure

Apple’s AI Strategy: Optimistic But Not Proven

The hosts come away cautiously optimistic about Apple’s direction, but not convinced it will execute well.

Main argument in Apple’s favor

  • Apple is no longer pretending it can build everything itself
  • Its strategy now seems more aligned with the actual state of AI
  • The company appears to understand that the phone itself may become the AI device

Main concern

  • Apple has a long history of promising more than it delivers on AI
  • Siri has repeatedly underperformed
  • The company’s success selling iPhones may have reduced the urgency to innovate

Notable takeaway

  • Apple’s AI challenge is less about vision and more about execution
  • The question is no longer “Does Apple have a plan?” but “Can Apple ship it?”

Anthropic’s Warning: Runaway AI Improvement and Safety Messaging

The second major topic is Anthropic’s claim that AI may be approaching a phase of recursive self-improvement, where models become powerful enough to improve themselves rapidly.

What Anthropic is arguing

  • AI coding systems are becoming dramatically more capable
  • Future models could potentially improve themselves or automate more of the engineering loop
  • That could create serious safety, security, and societal risks
  • Anthropic suggests it may be wise to slow frontier model development

The hosts’ reaction

  • Ranjan is highly skeptical and sees the messaging as repetitive and often performative
  • He argues the company has been saying some version of this for a long time
  • The continued push to raise capital, grow aggressively, and prepare for an IPO makes the “slow down” message feel inconsistent
  • He sees the company’s safety posture as partly genuine and partly PR strategy

Key tension discussed

  • If the technology is truly dangerous, why keep racing ahead?
  • One argument is that Anthropic must stay at the frontier if it wants to shape how the technology is deployed
  • Another argument is that the company is simply behaving like a normal high-growth business, despite safety rhetoric

Important insight

  • The most plausible explanation may be that Anthropic contains multiple internal camps:
    • marketing and business people
    • core researchers
    • true safety believers
  • These groups may not fully agree, which creates mixed signals in public messaging

Microsoft vs. OpenAI: From Partners to Competitors

The final major topic is Microsoft’s increasingly direct competition with OpenAI.

What Microsoft is signaling

  • At its Build event, Microsoft emphasized that it is no longer fully reliant on OpenAI
  • The company wants to be seen as a serious model builder in its own right
  • Mustafa Suleyman framed Microsoft as aspiring to become one of the top four AI labs
  • The company is leaning into a broader AI platform strategy across Copilot, Office, Azure, and other products

Why this matters

  • Microsoft’s relationship with OpenAI now looks more like a competitive alliance than a partnership
  • OpenAI may eventually want to build its own cloud infrastructure, which would threaten Azure
  • That possibility may be pushing Microsoft to develop its own model stack more aggressively

The hosts’ take

  • Microsoft likely has the products and distribution to matter in AI
  • But the strategic logic of building frontier models is less obvious than for companies like OpenAI or Google
  • A lot of Microsoft’s recent product naming and positioning is seen as overly complicated or awkward
  • Still, the company has a real chance if model routing and AI product specialization become the norm

Main Takeaways

  • Apple may finally be taking a practical, distribution-first approach to AI, with Siri at the center.
  • Execution remains the big question for Apple; the vision is plausible, but the company has not earned trust yet.
  • Anthropic’s safety warnings may be partly sincere, but they also serve branding and competitive positioning.
  • Microsoft and OpenAI are increasingly rivals, and the real strategic issue may be cloud infrastructure as much as model quality.
  • Across all three stories, the common theme is that AI is shifting from pure model hype to product, platform, and deployment strategy.

Notable Themes and Commentary

On Apple

  • Apple’s strength is its control of the interface layer
  • The iPhone may itself become the most important AI device
  • Siri’s future depends on whether Apple can make AI actually useful inside the OS

On Anthropic

  • Safety messaging is powerful, but it only works if it matches behavior
  • A company can be both serious about risk and still aggressively competitive
  • Public-market pressures may complicate any genuine slowdown effort

On Microsoft

  • Distribution remains a major advantage
  • Azure and the enterprise ecosystem give Microsoft a long runway
  • The company may be preparing for a world where it cannot rely on OpenAI forever

Final Verdict

This episode argues that the AI race is moving into a new phase: less about who has the flashiest demo, and more about who can embed AI into the products people already use. Apple may finally be moving in that direction, Anthropic is trying to reconcile safety and growth, and Microsoft is preparing for a post-exclusive OpenAI relationship.