Apple's best product ever

What It Covers

The Vergecast hosts David Pierce and Nilay Patel rank the 50 best Apple products of all time using 1.6 million audience votes, while also analyzing OpenAI's $122 billion funding round, the pivot of AI companies toward enterprise software, Mustafa Suleiman's redefinition of "superintelligence," and OpenAI's acquisition of podcast network TBPN.

Key Questions Answered

• •AI Industry Pivot: Every major AI company is quietly shifting from consumer-facing products toward enterprise and B2B software, where actual revenue exists. OpenAI killing Sora in favor of Codex, Microsoft reframing superintelligence as "delivering product value for enterprises," and the broader retreat from creative AI tools all signal the same directional move. The gap between what was promised to consumers and what is being built is widening rapidly and unavoidably.
• •Marketing Cannot Fix a Product Problem: OpenAI's acquisition of TBPN, a VC-adjacent tech livestream network, reflects a belief that AI's poor public perception is a communications failure. Polling shows 55% of people believe AI is not going well. With 900 million weekly ChatGPT users already forming direct opinions from firsthand product experience, no amount of branded media content can override what users already know from using the product themselves.
• •Enterprise vs. Consumer AI Gap: The only AI consumer product with genuine mass-market potential would resemble true AGI — an always-available, autonomous assistant. Everything currently shipping falls short of that bar. Automating Excel and business logic is genuinely valuable and commercially sound, but it is categorically different from the transformative personal computing revolution AI companies have been publicly promising for three years, creating a credibility collision that is accelerating.
• •Google's Structural Advantage in AI: Google can sustain losses on AI products like its VO video generation model indefinitely because search advertising generated roughly $100 billion last quarter. OpenAI, Anthropic, and other independents have no equivalent revenue cushion. This asymmetry means Google can treat AI as a subsidized internal project while competitors must find paying customers fast, making OpenAI's plan to compete directly with Google for consumer attention particularly precarious.
• •Apple Product Ranking Methodology: The Verge ran a head-to-head bracket-style ranker using a chess-based Elo scoring system across 50 Apple products, generating 1.6 million votes. The system surfaced strong recency bias — AirDrop, Apple Pay, and FaceTime all ranked in the top 20 — while foundational software like Mac OS X and QuickTime ranked far lower than their historical significance warrants. Audience familiarity with current products consistently outweighed historical impact in voting patterns.
• •Mac OS X as Apple's Most Consequential Product: Both hosts independently ranked Mac OS X as Apple's single most important product, though audience voting placed it outside the top five. The operating system, born from Apple's acquisition of NeXT, became the foundation for iOS, iPadOS, watchOS, and tvOS. Without it, none of Apple's post-2001 hardware successes exist. The audience's failure to recognize this stems from iOS users not knowing that iOS began as Mac OS X.
• •OpenAI's Structural Vulnerabilities in 2025: OpenAI enters the year having lost its tight Microsoft distribution advantage, with its Apple Intelligence partnership appearing to weaken as Apple opens Siri to multiple model providers. It must now compete against Anthropic in enterprise coding tools and against Google in consumer search — simultaneously. The $122 billion funding round, while large, comes from existing investors with circular financial interests, including Nvidia, which receives much of that capital back through chip purchases.