On Friday, February 6th, ABC News aired an interview with Daniela Amodei, co-founder of Anthropic—the artificial intelligence developer behind the Claude series of large language models, which rank only behind OpenAI’s ChatGPT in both popularity and performance. The wide-ranging discussion, which addressed Anthropic’s efforts to influence regulatory conversations and align their models with human values, was of particular interest to me: I’m currently writing a senior thesis on the way non-reductive approaches in the philosophy of mind can be used to deflate ascriptions of moral agency and consciousness to AI systems.
But another part of the interview stood out. Amodei, who studied English as an undergraduate, suggested that as a fast-growing company, Anthropic was especially interested in job candidates with the emotional intelligence and soft skills that are characteristic of a background in the academic humanities. Claude already serves as an extremely sophisticated coding assistant, which many consider stronger (and certainly faster) than the average entry-level programmer. But several of the current bottlenecks in AI development—alignment, reinforcement learning, and organized reasoning capacities—are not purely technical in nature. Creative thinking, problem solving, and qualitative judgment will be just as central to the AI revolution as are data centers, compute infrastructure, and memory capacity.
Amodei took the point even further:
“I actually think studying the humanities is going to be more important than ever…a lot of these models are actually very good at STEM. But I think this idea that there are things that make us uniquely human—understanding ourselves, understanding history, understanding what makes us tick—I think that will always be really, really important. And I think the ability to have critical thinking skills and learn how to interact with other people will be more important in the future, rather than less.”
It’s hard to tell how genuine comments like these can be. Part of Amodei’s job is to assuage the public’s concern that models like Claude will displace huge numbers of entry-level white collar workers, creating an unprecedented crisis of structural unemployment—something that her brother and Anthropic co-founder, Dario Amodei, readily acknowledged less than a year ago. For the Amodeis, the incentives here are mixed: on the one hand, hyping up the transformational impacts of AI can fuel investment in Anthropic and temporarily boost shareholder value; on the other, it can push the public debate on AI regulation in a direction that ultimately threatens the company’s bottom line.
Either way, Daniela Amodei is probably right that the widespread integration of AI into white-collar work will reconfigure the demand for human capital in a way that is unfavorable to those in STEM fields, especially software engineering and financial analysis. The most recent data suggests that the market for entry-level programmers is already in freefall. It remains unclear whether humanities degrees will appreciate in value as a result, or whether the loss will be absolute—although the latter seems more realistic.
One corollary of this moment of crisis for white-collar work is that the discourse surrounding ‘practicality’ and the humanities will soon adjust to new realities. The bogeyman of the unemployed English or philosophy major may soon become that of the college graduate in general. Or, if Amodei is right, the humanities degree may soon become the ‘practical’ or marketable option for college grads, with computer science, applied math, and economics pursued only for their edificatory merit but not their ‘utility’. Viewed purely from the standpoint of material interest, this might be good news for some graduates (like me), who have invested their time and energy in subjects (philosophy) with few applications beyond the academy. But it also raises broader questions about the purpose of a college education. Do we study Kant, Fanon, Ibn Tufayl and De Beauvoir; the medieval Mediterranean, 19th-century West Africa, and premodern China; Auden, Achebe, and Woolf; because it’ll help us get a job? Does the Core do that? No, at least not directly. The point, if there is one, is to broaden our moral and intellectual horizons, to become more attuned to the suffering of others, to better understand our place in the cosmos, to build a better world. The purpose of tertiary education, really, is in the name—to learn widely and grow substantially.
College should not be a place where students learn only to adapt to the strictures of American professional life. But if we accept the Amodei narrative, and approach the humanities as we do many (but certainly not all) STEM fields—as comprehensive skill development programs—then we risk treating them as means rather than ends in themselves, sapping them of what is singular and valuable to begin with.