Imagine a world where artificial intelligence is reshaping industries at lightning speed, yet a critical question looms: Can the United States keep pace in the manufacturing realm needed to fuel this technological revolution? That's the electrifying tension that defined 2025 for Stratechery, a publication that's now back on American soil after over a decade abroad—and trust me, it's got everyone talking. But here's where it gets really intriguing: Will AI be America's savior, or is it just highlighting deeper vulnerabilities? Let's dive in and unpack it all.
Back in 2013, I launched Stratechery while residing in the United States, but within just a few months, I relocated to Taiwan—check out that original post on independence for the full backstory. As a result, the bulk of my writing over the years has come from the opposite side of the globe compared to the tech giants I often analyzed. Fast-forward to this year, and here's the big news: I've made my way back to the U.S., marking Stratechery's 13th year as a domestically based outlet once more. For more on this personal shift and a quick break, see my update from earlier this year.
To give you a nostalgic recap, here are the links to my 'Taiwan Years in Review' for the past dozen years:
2024 | 2023 | 2022 | 2021 | 2020 | 2019 | 2018 | 2017 | 2016 | 2015 | 2014 | 2013
This year, the spotlight was overwhelmingly on artificial intelligence, especially within the realms of major tech firms. Yet, running counter to that progress was a growing worry about America's standing in manufacturing—a sector essential for turning innovative ideas into tangible products. It's fascinating, and maybe not too shocking, that these two themes have surged simultaneously. Could AI rescue the U.S. economy, and more importantly, can America produce the hardware required to truly harness AI's potential? And this is the part most people miss: the answer might hinge on bold policy decisions that could divide opinions nationwide.
In terms of output, Stratechery delivered 26 free articles (accessible in the articles category), 109 subscriber-exclusive updates (found in the daily email section), and 39 interviews (organized in the interviews category). What follows is a roundup of the majority of those articles, interviews, plus a curated selection of my top updates from the year, all designed to keep things engaging and informative.
Let's kick off with the five most popular articles, based on page views, which really captured readers' imaginations:
- DeepSeek FAQ – This piece demystifies DeepSeek, a game-changer that's flipped expectations for AI and global rivalry, particularly with China. For beginners, think of DeepSeek as an AI model that's outperforming rivals in efficiency and openness, sparking debates on whether it's a threat or an opportunity.
- Google, Nvidia, and OpenAI – Here, we explore how OpenAI and Nvidia face challenges from Google, but I argue OpenAI could come out on top with a solid advertising strategy to outmaneuver Google as a dominant 'aggregator'—a term for platforms that connect users, content, and services seamlessly.
- The Agentic Web and Original Sin – This delves into Microsoft's bold vision for an open agentic web, where AI agents operate autonomously. However, it highlights the need for integrated digital payments to build new marketplaces for AI-generated content, raising ethical questions about monetization and control.
- U.S. Intel – Diving into the government's potential equity stake in Intel, this article calls it a flawed but perhaps necessary step to revive Intel's foundry operations. For context, a foundry is like a semiconductor factory that produces chips for others, and this move could spark controversy over government involvement in private business—do you think this is smart intervention or overreach?
- The Benefits of Bubbles – Acknowledging we're in an AI bubble (a period of inflated hype and investment), it questions whether the resulting infrastructure and innovations will justify the frenzy. As an example, think of past tech bubbles like the dot-com era; they often lead to overinvestment, but also to breakthroughs that shape the future.
Shifting gears to a broader analysis of AI and its societal ripples—because nearly every piece touched on this topic in some way:
- AI’s Uneven Arrival: Building on advancements like o1/o3 models from OpenAI, this explores the path to AGI (Artificial General Intelligence, which can handle a wide range of human-like tasks). Yet, it warns that widespread adoption might lag, drawing parallels to how digital advertising, despite its potential, took years to dominate markets.
- Deep Research and Knowledge Value: Positioning Deep Research as a specialized AGI tool for niche areas, it argues that its ability to scour the entire internet will elevate the worth of exclusive knowledge, much like how proprietary data gives companies an edge in today's economy.
- Checking In on AI and the Big Five: A status update on AI through the perspectives of tech giants like Google, Apple, Microsoft, Amazon, and Meta, explaining how each is navigating the AI landscape for newcomers to follow.
- Tech Philosophy and AI Opportunity: This categorizes AI players by their philosophies and market strategies, clarifying who might win or lose in this evolving game—perfect for understanding why some startups thrive while others stumble.
- Content and Community: Contrasting traditional content models rooted in physical locations with emerging virtual communities, it suggests content as the glue for online groups, like how social media platforms have transformed how we connect globally.
Now, onto the big tech players—the ones pouring billions into AI and risking the most, plus insights into self-driving cars and entertainment's future:
- Facebook is Dead; Long Live Meta: Analyzing Meta's stellar earnings alongside Mark Zuckerberg's AI push, this suggests a synergy that's no accident. Plus, in a related piece on Sora, AI Bicycles, and Meta Disruption, it notes Sora's viral potential in creativity, which could uplift society but pose challenges for Meta's social network dominance.
- The YouTube Tip of the Google Spear: Celebrating Google's fluid structure, this highlights YouTube's AI prospects as a bullish indicator, explaining for beginners how AI could enhance video recommendations and creator tools.
- OpenAI’s Windows Play: Framing OpenAI as the 'Windows of AI'—a ubiquitous platform controlling hardware and software—this piece examines its strategy to become indispensable, akin to how Microsoft once dominated computing.
- Robotaxis and Suburbia: Predicting how autonomous taxis could bridge urban-suburban divides, potentially weakening cities and companies like Uber. Imagine robotaxis making suburban commutes as efficient as urban ones—exciting for convenience, but worrisome for city economies.
- Netflix and the Hollywood End Game: Positioning Netflix as a driver in Hollywood's transformation, betting on intellectual property value while countering threats like YouTube, with examples of how streaming has disrupted traditional studios.
Apple, meanwhile, got special attention this year after some missteps in AI integration:
- Apple AI’s Platform Pivot Potential: With AI delays, this advises Apple to focus on empowering developers for AI apps, rather than overextending, to maintain its ecosystem strength.
- Apple and the Ghosts of Companies Past: Noting Apple's uncertain long-term outlook for the first time in ages, it urges timely changes, drawing lessons from fallen giants like Kodak.
- Apple Retreats: Viewing Apple's WWDC as a strategic pullback, this praises it as a positive reset, even if it means retreating from ambitious plans.
- Paradigm Shifts and the Winner’s Curse: Explaining how past leaders struggle in shifts like AI, using the 'winner's curse' concept where success breeds complacency—do you agree Apple and Amazon are particularly vulnerable here?
- iPhones 17 and the Sugar Water Trap: Despite impressive new iPhones, this critiques Apple's growing detachment from world-changing trends, like how soda companies once ignored health shifts.
Turning to America's broader hurdles—leading in AI but lagging in manufacturing—these pieces probe policy implications:
- AI Promise and Chip Precariousness: Celebrating AI's vibrancy while stressing semiconductor shortages, it calls for urgent action, as chips are the 'brain' powering AI devices.
- American Disruption: Re-evaluating tariffs under Trump, this uses a 'disruption' lens to diagnose manufacturing woes, advocating demand-driven solutions over protective barriers to stimulate innovation.
- Resiliency and Scale: Arguing that cheaper transport and communication boost efficiency but erode resilience, it suggests embracing some inefficiency for true durability, like how over-reliance on global supply chains backfired during pandemics.
Stratechery Interviews kept the conversation lively, featuring weekly podcasts and transcripts (skipping July and August) with executives, founders, and experts:
Public Company Executives:
ServiceNow CEO Bill McDermott on enterprise AI agents | Uber CEO Dara Khosrowshahi on aggregation and autonomy | Snowflake CEO Sridhar Ramaswamy on data and AI | Google Cloud CEO Thomas Kurian on building enterprise culture | Meta CEO Mark Zuckerberg on AI and social media evolution | SAP CEO Christian Klein on enterprise AI | Nvidia CEO Jensen Huang on chip controls, AI factories, and enterprise pragmatism | Cloudflare CEO Matthew Prince on internet history and pay-per-crawl | YouTube CEO Neil Mohan on creating stages for creators | Booking CEO Glenn Fogel on travel aggregation | Asana Founder Dustin Moskovitz on AI, SaaS, and safety | Unity CEO Matthew Bromberg on turnarounds | Atlassian CEO Mike Cannon-Brookes on Atlassian and AI | Rivian CEO RJ Scaringe on car-building and autonomy
Startup Founders:
Anduril CEO Brian Schimpf on paradigm shifts | Manna CEO Bobby Healy on drone delivery | Tailscale CEO Avery Pennarun on networking innovations | OpenAI CEO Sam Altman (March and October) on consumer tech and DevDay | Plaid CEO Zach Perret on trust in finance | Cursor CEO Michael Truell on AI-assisted coding | Sierra CEO Bret Taylor on AI agents and tech history | Flighty CEO Ryan Jones on app development in 2025
Analysts:
Jon Yu on YouTube and semiconductors | Daniel Gross and Nat Friedman on AI models, margins, and moats | Matthew Ball on the gaming slump | Benedict Evans on AI unknowns | Michael Nathanson on streaming (March and December) | Dan Kim and Hassan Khan on CHIPS, and Dan Kim on Intel/Nvidia/government | Eric Seufert on digital advertising (April and November) | Patrick McGee on Apple in China | Ben Bajarin on AI infrastructure | Gracelin Baskaran on rare earths | Michael Morton on AI and e-commerce
Rounding out the year, here are some standout Stratechery Updates that stood out to me:
- January 8: Meta's moderation policy shifts, Zuckerberg's path—and mine, plus the boldness of strategic imitation.
- February 4: Apple's financials, OpenAI's Deep Research, and the breakdown of verification processes.
- February 19: Encryption's tricky balance, Netflix's results, and the snowballing benefits of aggregation.
- March 3: Microsoft's Skype shutdown, its origins, and Microsoft's charitable efforts.
- March 5: Alexa's evolution, a quick Alexa history, and Amazon/Apple's missteps.
- March 19: Nvidia's conference and ASICs, power limits, and performance trade-offs.
- April 15: Meta vs. FTC, Facebook's phases, video quality issues, and market dynamics.
- May 14: Airbnb's app revamp, experiences/services focus, and Chesky's leadership style.
- May 21: Google's event, search mechanics, and product expansions.
- June 3: Nike's Amazon venture, its risky shift, inevitability vs. intention, and Amazon's role.
- June 24: Talent competition, NBA finances, and AI funding.
- July 14: Google and Windsurf, questionable deals, tradition's role, and Silicon Valley's network.
- July 16: Cloudflare's content freedom day, Google's edge, and AI monetization.
- July 28: TSMC's performance, A16 chip details, backside power techniques, Intel's updates, architecture, and AI.
- July 30: Figma's IPO, the Figma OS concept, and its AI potential.
- August 27: K-pop's dark hunters, Sony's gambles, and Netflix's aggregation strategy.
- September 9 and November 11: SpaceX's spectrum acquisitions, specifics, big bets, pivots, and potential Apple-SpaceX partnerships.
- September 24: YouTube's account restorations, free speech debates, cultural norms, and platform influence.
- November 3: Google's and Meta's earnings, and Reality Labs' expenses.
- December 10: Trump's green light for H200 chip exports to China, the graduated approach, and if it was a wise call.
I'm deeply thankful to all the subscribers who enable this passion to be my profession. Wishing you a joyful Christmas and a prosperous New Year—can't wait for what 2026 brings!
As we wrap up, let's spark some debate: Is government intervention in tech, like the Intel stake or Trump's chip decision, a necessary evil for national competitiveness, or does it stifle innovation? And on a broader scale, do you believe the AI bubble will burst with lasting benefits, or is it setting us up for disappointment? Share your thoughts in the comments—do you agree, disagree, or see a middle ground?
