Microsoft co-founder and philanthropist Bill Gates ’77, LL.D. ’07, famously dropped out of Harvard in 1975 to focus on building a new kind of software company. Fifty years later, this February 3, he returned to campus and discussed the newest innovation in technology—artificial intelligence.
Gates sat down with Montgomery professor of the practice of public leadership Arthur Brooks in Sanders Theatre for a discussion about his memoir, Source Code, publishing this month. For Gates, computing has always been about pushing boundaries and democratizing access. His early years at Microsoft brought computing—once the exclusive realm of massive, centralized machines accessible only to governments and corporations—into the hands of individuals.
The conversation naturally turned to the next frontier of technology. During the question period (a reprise of an earlier such session on campus), students seemingly asked a variant of the same question: Okay, but what about AI?
Gates framed the current shift as a continuation rather than a departure from past evolutions in computing technology: “What we’re doing now [in artificial intelligence],” he explained, “is kind of an extension of the digital revolution—because it’s built on the chips and the internet, and all of those things are about free intelligence.”
While early computing was constrained by hardware limitations and a lack of software sophistication, AI today is primarily limited by training data and interpretability. These are barriers that, like those in the past, Gates believes will soon be overcome. Yet AI also diverges sharply from the personal computing revolution. Unlike the first PCs, which merely amplified human capabilities, AI has the potential to replace them. Gates pointed out the existential shift: “Intelligence will be completely free.” What does that mean? Computing was once about making existing tasks more efficient, he explained, but AI could fundamentally redefine what tasks humans delegate to people or to machines.
However, this development of “free intelligence” isn’t all positive. Gates once believed that the “digital divide” was primarily about access: “Getting computers into inner-city schools,” he said, and other underserved communities. Today, though, he sees it differently. “[Sometimes when] you empower humans, it doesn’t always get pushed in the right direction.” The problem isn’t merely access but ensuring there isn’t an abuse of access.
For instance, Gates noted that one of the “sins of computing” is its ability to reinforce false narratives. AI, he acknowledged, could either exacerbate or mitigate this problem, depending on how it is deployed. Now technology influences relationships and the spread of knowledge, allowing misinformation and extreme ideologies to flourish (read more in “When Technology and Society Clash,” November-December 2024).
On Artificial Intelligence in Education and Medicine
Artificial intelligence, Gates argued, could fundamentally change classroom dynamics. Imagine an AI tutor that not only understands pedagogy, he suggested, “but also motivation.” These systems wouldn’t just deliver lessons; they would measure student engagement, identify struggles, and adjust in real time, so that no one gets left behind (read more in “Applying AI—How and Why,” March-April 2024).
In higher education, AI could revolutionize research and administrative processes, helping students navigate complex coursework and institutions tailor curricula to evolving industries. Yet, while AI could eventually provide scalable, individualized learning, it can’t replicate human connection, mentorship, empathy, and the ethos of a wise teacher. Brooks repeatedly turned to a theme throughout the discussion: Gates’s journey to success was at its deepest level “about relationships, not technology,” in Brooks’ estimation.
Gates also spoke with optimism about AI’s potential in medicine. “It’s very profound and even a little bit scary—because it’s happening very quickly, and there is no upper bound,” he said. One of its most immediate and transformative applications, he suggested, is in medical diagnosis. Healthcare is on the brink of transformation from AI—from use cases in identifying rare diseases to customizing treatments based on genetic profiles (read more in “Predicting Viral Variants and Vaccine Cures,” November-December 2024, and “AI as a Cancer Oracle?” May-June 2024). Unlike human doctors, whose diagnostic abilities are limited by experience and bandwidth, AI can synthesize vast amounts of medical literature, patient histories, and emerging research in real time.
Gates envisions AI-powered tools that can provide primary-care diagnostics, reducing dependence on overburdened medical professionals. “Eventually,” he noted, there will be no shortage of doctors, “and the machine will probably be superior to humans—because the breadth of knowledge that you need to make some of these decisions really goes beyond individual human cognition.” This shift, he argued, is not merely about efficiency; it’s about equity. In developing nations, where doctor-to-patient ratios remain abysmally low, AI could bridge the gap, bringing world-class diagnostic capabilities to those who need it most.