The left side of figure 5-1 indicates that when there is a performance gap — when product functionality and reliability are not yet good enough to address the needs of customers in a given tier of the market — companies must compete by making the best possible products. In the race to do this, firms that build their products around proprietary, interdependent architectures enjoy an important competitive advantage against competitors whose product architectures are modular, because the standardization inherent in modularity takes too many degrees of design freedom away from engineers, and they cannot not optimize performance.

The issue I have with this analysis of vertical integration — and this is exactly what I was taught at business school — is that the only considered costs are financial. But there are other, more difficult to quantify costs. Modularization incurs costs in the design and experience of using products that cannot be overcome, yet cannot be measured. Business buyers — and the analysts who study them — simply ignore them, but consumers don’t. Some consumers inherently know and value quality, look-and-feel, and attention to detail, and are willing to pay a premium that far exceeds the financial costs of being vertically integrated.

Google trains and runs its Gemini family of models on its own TPU processors, which are only available on Google’s cloud infrastructure. Developers can access Gemini through Vertex AI, Google’s fully-managed AI development platform; and, to the extent Vertex AI is similar to Google’s internal development environment, that is the platform on which Google is building its own consumer-facing AI apps. It’s all Google, from top-to-bottom, and there is evidence that this integration is paying off: Gemini 1.5’s industry leading 2 million token context window almost certainly required joint innovation between Google’s infrastructure team and its model-building team.

On the other extreme is AWS, which doesn’t have any of its own models; instead its focus has been on its Bedrock managed development platform, which lets you use any model. Amazon’s other focus has been on developing its own chips, although the vast majority of its AI business runs on Nvidia GPUs.

Microsoft is in the middle, thanks to its close ties to OpenAI and its models. The company added Azure Models-as-a-Service last year, but its primary focus for both external customers and its own internal apps has been building on top of OpenAI’s GPT family of models; Microsoft has also launched its own chip for inference, but the vast majority of its workloads run on Nvidia.

Google is certainly building products for the consumer market, but those products are not devices; they are Internet services. And, as you might have noticed, the historical discussion didn’t really mention the Internet. Both Google and Meta, the two biggest winners of the Internet epoch, built their services on commodity hardware. Granted, those services scaled thanks to the deep infrastructure work undertaken by both companies, but even there Google’s more customized approach has been at least rivaled by Meta’s more open approach. What is notable is that both companies are integrating their models and their apps, as is OpenAI with ChatGPT.

It may be the case that selling hardware, which has to be perfect every year to justify a significant outlay of money by consumers, provides a much better incentive structure for maintaining excellence and execution than does being an Aggregator that users access for free.

Google’s collection of moonshots — from Waymo to Google Fiber to Nest to Project Wing to Verily to Project Loon (and the list goes on) — have mostly been science projects that have, for the most part, served to divert profits from Google Search away from shareholders. Waymo is probably the most interesting, but even if it succeeds, it is ultimately a car service rather far afield from Google’s mission statement “to organize the world’s information and make it universally accessible and useful.”

The only thing that drives meaningful shifts in platform marketshare are paradigm shifts, and while I doubt the v1 version of Pixie [Google’s rumored Pixel-only AI assistant] would be good enough to drive switching from iPhone users, there is at least a path to where it does exactly that.

the fact that Google is being mocked mercilessly for messed-up AI answers gets at why consumer-facing AI may be disruptive for the company: the reason why incumbents find it hard to respond to disruptive technologies is because they are, at least at the beginning, not good enough for the incumbent’s core offering. Time will tell if this gives more fuel to a shift in smartphone strategies, or makes the company more reticent.

while I was very impressed with Google’s enterprise pitch, which benefits from its integration with Google’s infrastructure without all of the overhead of potentially disrupting the company’s existing products, it’s going to be a heavy lift to overcome data gravity, i.e. the fact that many enterprise customers will simply find it easier to use AI services on the same clouds where they already store their data (Google does, of course, also support non-Gemini models and Nvidia GPUs for enterprise customers). To the extent Google wins in enterprise it may be by capturing the next generation of startups that are AI first and, by definition, data light; a new company has the freedom to base its decision on infrastructure and integration.

Amazon is certainly hoping that argument is correct: the company is operating as if everything in the AI value chain is modular and ultimately a commodity, which insinuates that it believes that data gravity will matter most. What is difficult to separate is to what extent this is the correct interpretation of the strategic landscape versus a convenient interpretation of the facts that happens to perfectly align with Amazon’s strengths and weaknesses, including infrastructure that is heavily optimized for commodity workloads.

Meta’s open source approach to Llama: the company is focused on products, which do benefit from integration, but there are also benefits that come from widespread usage, particularly in terms of optimization and complementary software. Open source accrues those benefits without imposing any incentives that detract from Meta’s product efforts (and don’t forget that Meta is receiving some portion of revenue from hyperscalers serving Llama models).

The iPhone maker, like Amazon, appears to be betting that AI will be a feature or an app; like Amazon, it’s not clear to what extent this is strategic foresight versus motivated reasoning.

achieving something approaching AGI, whatever that means, will require maximizing every efficiency and optimization, which rewards the integrated approach.

the most value will be derived from building platforms that treat models like processors, delivering performance improvements to developers who never need to know what is going on under the hood.

Upon arriving at the facility, Bruce handed in his smartphone and passed through metal detectors. He was in awe of the semiconductor production line: Overhead rails carried wafers from one station to another while workers in white protective suits kept the machinery running. “It really just felt like I was touring some kind of living thing that was greater than humans; that was bigger than us,” Bruce recalled.

TSMC made attempts to bridge some of the cultural differences. After the American trainees asked to contact families and to listen to music at work, TSMC loosened the firewall on T phones to allow all staff access to Instagram, YouTube, and Spotify. Some Taiwanese workers attended a class on U.S. culture, where they learned that Americans responded better to encouragement rather than criticism, according to an engineer who attended the session.

Several former American employees said they were not against working longer hours, but only if the tasks were meaningful. “I’d ask my manager ‘What’s your top priority,’ he’d always say ‘Everything is a priority,’” said another ex-TSMC engineer. “So, so, so, many times I would work overtime getting stuff done only to find out it wasn’t needed.”

Training in Taiwan, which typically lasted one to two years, wasn’t all miserable, the Americans said. On the weekends, the trainees traveled across the island, marveling at the country’s highly efficient public transport network. Bruce spent his weekends hiking and frequenting nightclubs. He chatted with the families that run night-market food stalls, and entertained strangers who requested selfies with foreigners.

For the Taiwanese, many of whom planned for extended stays in Phoenix, that meant relocating entire families — toddlers and dogs included — to a foreign country. Many regarded it as a once-in-a-lifetime opportunity to explore the world, practice English, and send their children to American schools. Younger families planned pregnancies so they could give birth to American citizens. “If we are going to have children, of course we will have them here,” a Taiwanese engineer told Rest of World. “As an American citizen, they will have more options than others.”

Many experienced a culture shock. The bustling cities of Taiwan are densely packed and offer extensive public transport, ubiquitous street food, and 24-hour convenience stores every few blocks. In northern Phoenix, everyday life is impossible without a car, and East Asian faces are scarce

“Everything is so big in America,” said one engineer, recalling his first impression. He recounted his wife summarizing her impression of the U.S.: “Great mountains, great rivers, and great boredom.”

Having spent years under the company’s grueling management, they were used to long days, out-of-hours calls, and harsh treatment from their managers. In Taiwan, the pay and prestige were worth it, they told Rest of World — despite the challenges, many felt proud working for the island’s most prominent firm. It was the best job they could hope for.

Sometimes, the engineers said, staff would manipulate data from testing tools or wafers to please managers who had seemingly impossible expectations.

A former TSMC staffer who worked on the education program said managers were instructed not to yell at employees in public, or threaten to fire them without consulting human resources. “They would say, ‘Okay, okay, I get it. I’m not going to do that,’” the employee recalled to Rest of World. “But I think in the heat of the moment, they forgot, and they did do it.”

Chang-Tai Hsieh, an economics professor at the University of Chicago, told Rest of World that TSMC had found the U.S. a challenging environment to operate in because of the complicated regulatory process, strong construction unions, and a workforce less used to the long hours that are commonplace at TSMC in Taiwan.

Sitting in a room together, the engineers admitted that although they had made some progress in acclimating to life in the U.S., TSMC had yet to find a balance between the two work cultures. Some Taiwanese workers complained that management was being too accommodating in giving Americans less work, paying them high salaries, and letting them get off work early.

Apple’s M-series silicon team is simply on fire, doing some of the most impressive work in the history of computer architecture design and engineering. PC laptops in this class weigh over 6.5 pounds and offer terrible battery life; the 16-inch MacBook Pro weighs 4.7–4.8, the 14-inch MacBook Pro just 3.4–3.6, and all offer remarkably long battery life. It’s not like Apple’s silicon team had one breakthrough moment back in 2020 and have since been regressing to the mean — they continue to increase their lead over the rest of the industry in performance-per-watt.