Careless People: A Cautionary Tale of Power, Greed, and Lost Idealism by Sarah Wynn-Williams. Macmillan, 2025.
Apple in China: The Capture of the World's Greatest Company by Patrick McGee. Simon and Schuster, 2025.
Empire of AI: Dreams and Nightmares in Sam Altman's OpenAI by Karen Hao. Penguin Random House, 2025.
The debate about whether universities should be run like businesses has reached the point of a cliché. In fact, it may no longer make sense to call it a debate, as the idea that universities should run more efficiently and pivot more nimbly to meet the needs of “the customers” (students) and their future employers has come to dominate the decision-making of private and public university administrations alike. Part of that effort has involved reducing the ranks of full-time, tenure-line faculty and instead relying on contingent labor, a long-term trend that has accelerated since the 2008 economic downturn and the onset of the COVID-19 pandemic. At the same time, administrations have worked to diminish the role of the faculty in decision-making processes. Academic institutions in the United States and beyond have increasingly invested in real estate and in technologies like large language models (LLMs). The idea has been to create a more agile and adaptable university that could, for example, more swiftly meet the challenges of educating students amid a global pandemic. Meanwhile, the results from these efforts have been mixed and have failed to insulate universities from declining enrollments and government attacks. The recent publication of three books about what are arguably among the most successful businesses in the twenty-first century led me to revisit questions about the relationship between higher education and the tech industry. What might universities learn from businesses, and specifically big tech, in the current climate? And what, if anything, might these companies learn from academia?
The three books provide a behind-the-scenes look at Facebook (Meta), Apple, and OpenAI, the first through a memoir by Sarah Wynn-Williams, former director of public policy at Facebook, and the other two through the lens of investigative journalists. Wynn-Williams’s at times gossipy, at times heart-rending account, Careless People, attempts to give the reader a sense of the broader culture and priorities at Facebook from her specific vantage point as a top policy official. Meta has since secured a ruling preventing the author from further publicizing the book, because she was in violation of a nondisparagement severance agreement. Patrick McGee (Apple in China) and Karen Hao (Empire of AI) both combine a panoramic view of the companies and their history with the details from former contractors and employees that offer nuance and specificity. All three authors provide a behind-the-scenes view of the tech industry, whether it involves decisions made by executives at headquarters, the engineers and coders working to make the technology accessible to consumers, or the content moderators and communities affected by the construction of new data centers across the globe. All three books help us to understand both how these companies achieved global ascendance and the tremendous harms they committed along the way—to individuals, to communities, to countries, and arguably to the entire planet.
While the three companies have more similarities than differences, a few of those differences are important to note. Apple emerged as a late-twentieth-century company that makes things. It hired executives from other giants of twentieth-century US manufacturing and developed a product line that relies on a supply chain in a more traditional sense. Even Steve Jobs, the former visionary leader of Apple, was interested in products and design—the tactile and aesthetic experience of using a phone, for example. On the other hand, Mark Zuckerberg and Sam Altman, of Meta and OpenAI, respectively, represent a departure from that tradition. Both executives famously dropped out of college and emerged as leaders who were skilled at promoting themselves and their ideas. Their businesses are, on a basic level, about facilitating communication and the pursuit of knowledge. But they use their platforms to leverage power and influence. According to Wynn-Williams and Hao, both Zuckerberg and Altman have entertained the idea of running for president of the United States, but now their sights seem set on using their influence to attain an even more expansive vision of global power.
All three companies intersect with the business of higher education. Apple, for instance, achieved a major breakout into the individual computing market when the original iMac, released in 1998 in an array of bright, translucent colors, became a highly desired product among college students. Today, Apple products are ubiquitous on college campuses among students, faculty, and staff alike. Facebook, with sinister 2004 origins as a “hot-or-not” ranking page for college students, remains an important platform for sharing information about campus groups and events, as does Instagram, the photo- and video-sharing platform Facebook acquired in 2012. And OpenAI changed the landscape of higher education with the launch of the LLM chatbot ChatGPT in 2022. Students were suddenly able to outsource the work of writing papers and other assignments with the click of a virtual button. What, beyond these obvious intersections, might those of us in higher education learn from these books?
One of the more significant takeaways from Apple in China and Empire of AI is that investments in research and development are central to the technological advances that enabled each company’s meteoric rise; research for research’s sake, which mostly happens in the context of academia, was essential to these breakthroughs. Initially, Apple relied heavily on engineers and designers educated in the United States, who then trained and supervised their counterparts working at various points in the supply chain. These US executives often operated with stereotypical assumptions—for instance, it was a widely held view within Apple that Chinese employees were capable only of imitating, not of innovating (more on this below). What they came to find, however, was that the Chinese government invested in its own research and training to the extent that, during the pandemic, it became evident that US engineers no longer needed to travel to China on a regular basis to train their Chinese counterparts. China achieved a highly capitalized version of import-substitution industrialization by investing in their workers and infrastructure, thereby reducing their reliance on engineers and designers from Apple and on suppliers from Taiwan and other countries.
OpenAI, a company that began with an altruistic idea that its technology would be accessible for the common good, teaches us a different lesson about research. As most university administrators will tell you, doing things for the common good can be quite expensive. In Empire of AI, Hao reveals that as Altman increasingly felt pressured to provide a return to investors, the “experiment in idealistic governance” unraveled. When Altman and the board made a controversial decision to switch course and shift from the open nonprofit model to a private for-profit model, the efforts to develop artificial intelligence became a race to be first rather than a democratic, transparent effort. According to Hao, this accelerated the research and development process, leaving little time to understand or interrogate flaws or potential downsides to the technology.
Moreover, in the race to be first, the line between academic research and industry blurred as industry recruited researchers to dual appointments and then poached them from university labs altogether. This was an easy feat because private companies could offer enormous salaries and access to far better equipment and computing capacity than universities could provide. The effect of this shift was to narrow the ideas that researchers were pursuing and to limit the knowledge-sharing that occurs in a more traditional academic environment. For example, whereas in the academic realm researchers need to publish to secure promotions and grants, in industry there is a premium on secrecy. As Hao put it, there was an “industry-wide shift from peer-reviewed to PR reviewed.” The lesson here is that research is expensive, and although universities may never be able to compete with the tech world in terms of salaries and benefits, more investments are needed to provide an environment that will attract researchers who are interested in working for the greater good. And while these books are specific to the tech world, there is a case to be made for increasing investments in research across the board.
One argument in favor of investment in research beyond the STEM fields is that all three books reveal a stunning lack of cultural and political awareness as each company extended its reach across the globe. In Empire of AI, Hao describes a meeting between American executives and community members about a new data center in Chile, where most of the executives did not speak Spanish and the lawyer who did seemed to mistranslate what the locals were saying, contributing to mistrust between the parties. Similarly, McGee observes in Apple in China that executives made a series of miscalculations about the abilities and intentions of their Chinese counterparts and Chinese government officials. A few key interlocutors like John Ford and Doug Guthrie, employees who did develop a deep knowledge about Chinese culture and society, prove to be the exceptions rather than the rule. These missteps and blunders are easily avoidable through investment in and utilization of the cultural and political knowledge and expertise found in universities across the country.
Wynn-Williams claims that she willed her position into existence when she observed the lack of thought and preparation that went into Facebook’s efforts to expand globally. She describes how she often prepared briefings on the leaders and top issues of various countries before Zuckerberg and other executives met with them, but those briefings were frequently ignored. In her telling, her warnings about the risks of working with authoritarian governments in Myanmar and China went unheeded. Specifically, she was concerned about how Facebook might be complicit in government censorship or might put dissidents at risk. Apple executives recognized that even as Chinese leader Xi Jinping became more authoritarian, the government became more predictable in a useful way. Under Xi, Apple executives felt that they better understood the priorities of the government and the terms under which they were expected to operate. The company had become so dependent on China for the supply chain and as a market that executives were willing to look the other way when it came to political freedoms and human rights abuses.
The extractive and exploitative practices of these companies, especially when it comes to employee relations, ought to give university administrators pause when they consider models to emulate. Karen Hao explains that she picked the title Empire of AI to draw attention to the extractive nature of generative AI and, specifically, LLMs. The central argument of the book is that OpenAI and other tech companies operate like empires: They expand, extract, and exploit, taking a human and environmental toll in the process. Her book outlines the multiple layers of extraction and exploitation, from the reliance on underpaid contract workers to perform essential tasks (for example, reviewing and removing illegal or harmful content) to the unrelenting quest for more data centers. She compares data collection to a new imperial land grab: Companies “mine” your data and then figure out how to sell it back to consumers.
These books serve as cautionary tales about the risks of that logic. The amount of human labor required to “teach” LLMs is astounding and shows that we cannot outsource teaching, despite the fantasies of the tech industry and some university administrators. But the tech industry has devalued everyone’s labor, converting skilled, educated workers into “ventriloquists” for chatbots and increasingly relying on subcontracted labor to obscure the relationship between workers and the companies. Beyond the devaluing of certain labor, these industries all extract an irreparable toll on their employees, whether it is the psychological damage to content moderators who are required to look at violent, pornographic images, the financial precarity experienced by employees as subcontractors decide to move on to a new country with even more desperate workers, or the health and emotional toll experienced by so many top executives. Wynn-Williams describes how Facebook demanded that she continue to work even while she remained on bedrest after nearly dying during childbirth and how the demands of her job strained her marriage. McGee reveals that many Apple employees died in their forties, and that even Steve Jobs blamed the stress of his job for his cancer. The travel demands on employees at Apple caused so many marital problems that standard procedures were in place to offer bonuses or extra time off for those facing a divorce. One of the areas where academia has consistently followed the path of the corporate world is to increasingly undervalue employees and treat them as disposable.
But what about the supposed inevitability of these technologies? Can we really escape the cycle of extraction and abuse that these books portend? On this point, the authors disagree. Wynn-Williams and McGee both suggest we are too far down the rabbit hole. McGee asserts that “Apple wasn’t interested in best practices, it wanted to upend how things were done,” similar to the now infamous Facebook motto, “Move fast and break things.” Hao argues there is another way forward, that “we don’t have to accept the logic of unprecedented scale and consumption.” As activist and organizer Kelly Hayes writes, “Not every harm or manipulation has been intentional, because Big Tech doesn’t move with intention. It moves with indifference, clawing its way to higher valuations and a larger user base.” Here she suggests what could be a way forward for academia—to move with intention to avoid replicating the harms. We can insist on best practices within our academic institutions, even and especially when it comes to the use of LLMs. Hao argues that AI can be “community driven, consensual, and respectful of local people and history” and that governance can be “inclusive and democratic.” These are values shared by many of us who work at universities, values I worry that administrators and governing boards are content to leave behind in the quest for efficiency or compliance with political pronouncements. Hao cites as an example the use of small, targeted AI tools designed for specific tasks like reviving the Māori language te reo. These tools require relatively little computational power, are free and accessible, and use contained training data, unlike the LLMs that consume so much energy and increasingly scrape the internet for all kinds of data, including data that may be unreliable or plagiarized. Again, researchers who can do independent evaluations outside of the companies can explore more efficient and targeted uses of AI beyond LLMs. These uses may not provide an immediate return to investors, but they can do important restorative work in our communities.
Any of these books would make for interesting case studies to discuss in business classes, but academics more broadly can learn a lot from the stories of companies that have such an outsized influence on all our lives. Beyond that, I can envision many other uses for these books in the classroom. As someone who studies Latin American history, I can imagine using chapters of Empire of AI to discuss both the labor and the environmental impacts of the tech industry on Latin American countries, and scholars of African cultures and societies will also find useful sections to do the same. Apple in China obviously would be of interest to scholars of China, but it also helps us understand commodity chains that extend beyond China to other parts of Asia and the wider world. Careless People urges us to consider the diplomatic implications of the tech industry and the roles these companies play in shaping our politics at the community and national levels. Indeed, we in higher ed can learn much from these books, and the most important lesson might be not to fall victim to the mantra that we must change for the sake of change rather than act with intention and in alignment with our values.
Lisa Pinley Covert is professor of history and past president of the AAUP chapter at the College of Charleston.