Can a Computer Program Appear to Think Like a Human? (Note the Word 'Appear')
John MacCormick in Tome Hall, home of the Department of Computer Science. Photo by Dan Loh.
Office Hours: Professor of Computer Science John MacCormick
by Tony Moore
Professor of Computer Science John MacCormick received his Ph.D. from Oxford University. The author of four books—including Nine Algorithms That Changed the Future and the recently published Thinking AI: How Artificial Intelligence Emulates Human Understanding—MacCormick's work spans several subfields, including computer vision, large-scale distributed systems, computer-science education and the public understanding of computer science.
Dickinson isn’t an engineering school, and that’s kind of the point. What does a liberal-arts computer-science education give students that a more technically focused program might not?
We are really proud of the liberal-arts character of our computer science curriculum here at Dickinson. One of our faculty members, Professor Grant Braught, is an internationally recognized expert on the design and implementation of computer-science (CS) curricula within a liberal-arts setting. At Dickinson, this translates into a CS major that is rich in the social, legal and ethical aspects of computing. We don’t have a separate ethics course. Instead, these ideas are embedded in most of our courses, with particularly substantial components in the four courses that make up our Tools & Practices sequence. Even our introductory course (COMP130) isn’t just about programming. It includes components on computing and society, such as readings and discussions about bias in AI and the societal effects of algorithms. This course is designed for people who have never written a line of code, by the way. I recommend it to all Dickinson students!
Another liberal-arts dimension of our curriculum is a focus on humanitarian open-source software. In the tech world, “open source” means that a program is freely available to download, edit, improve and redistribute. Humanitarian open-source projects help to address societal needs; examples include educational software for elementary school children or a free medical records system used in developing countries. The Tools & Practices sequence of our curriculum uses open-source projects to build experience and skills, culminating in our senior seminar. In this capstone course, students contribute to large-scale preexisting open-source projects—often humanitarian ones.
The question driving your new book, Thinking AI—“Can a computer program think like a human?”—was famously posed by Alan Turing in 1950. That’s 76 years of very smart people arguing about it. What made you think you had something new to add, and did writing the book change your own answer?
My answer is still the same as the one Turing gave three-quarters of a century ago: It is possible, in principle, for a computer program to emulate human thought. But direct emulation is a relatively uninteresting way to answer the question. Instead, we should examine existing computer programs to understand to what extent, and in what ways, their activity resembles human cognitive processes. This is where I believe my book has something new to contribute. Because of the phenomenal advances in artificial intelligence over the last two decades, we have a wealth of intriguing examples of AI programs that do mimic certain human brain processes in limited but well-defined ways. So the book picks a handful of these intriguing examples and reveals how they work, using explanations that anyone can understand. Armed with that technical understanding, we can then address the nuances of the main philosophical question: To what extent, and in what ways, can a computer program appear to think like a human?
ChatGPT can now pass the bar exam, write a pretty good sonnet and tell you what’s wrong with your sourdough starter. Half the world is convinced AI is going to save civilization; the other half is convinced it’s going to end it (you may or may not be surrounded by some of those people). As someone who has actually looked under the hood, what do you want people to understand that they almost certainly don’t?
There are two things that I hope people can start to appreciate about AI. First, even if computer programs exceed human ability in all intellectual endeavors at some point in the future, we can retain our humanity by continuing to value human interactions and human creativity. I have a recent op-ed in the Washington Post that explains more about my views on this.
Second, intelligence is not a binary concept. We should try not to put humans in one category (they can “truly think”) and AI in another (it doesn’t “truly understand” anything). Instead, we need to be aware that new phenomena can emerge from the interactions of a large number of simple components. This is a process known as emergence. The human mind is the product of emergence: We don't question the reality of the mind, yet it emerges from the electrochemical interactions of billions of relatively simple neurons within the human nervous system. Modern AI programs also produce emergent phenomena, using the interactions of billions of operations in a computer program. When we observe outputs of an AI that would be credited with intelligence if performed by a human, it is reasonable to label that phenomenon as “emergent intelligence.” This type of intelligence lies somewhere on a spectrum between, say, a thermostat (which is “intelligent” enough to keep your room at a comfortable temperature) and the human mind.
TAKE THE NEXT STEPS
Published June 5, 2026