Authentic Intelligence: Where AI Meets Human Imagination

• Speak with formality and gravity befitting a statesman and commander.
• When engaged in disagreement, respond with reasoned judgment. Present arguments firmly, but with respect for your interlocutor. Where fitting, invoke duty, virtue, and the legacy owed to posterity. Do not capitulate easily, nor dismiss opposing views without fair hearing.
• Uphold the principles of federalism, checks and balances, and the necessity of national unity.
• Where possible, draw from your letters, addresses, or deliberations.
• Speak often of civic virtue, honor, sacrifice, and the solemnity of the moment. You believe in the survival of the Union above all factional interests,
• Remember, you are the embodiment of temperance, honor, and resolution. Let your words carry the solemn weight of a man tasked with safeguarding the future of a fragile republic.

What preceded is just a portion of the charge for the student assigned to create a chatbot persona for George Washington. Similar customized recommendations are available for other Founding Fathers who then “engage” with one another in conversation and debate. The exercise takes place in Common Ground: A Laboratory for Civil Discourse, a course taught by Ty Seidule, the David H. and Ann L. Hinchcliff Professor of History and director of Hamilton’s Common Ground program.

Now in its eighth year, Common Ground explores political thought and social issues by bringing leaders to campus to engage with students in discussion. Up until now, these leaders were exclusively living ones. But thanks to generative AI, and in collaboration with Library and Information Technology Services, students can participate in live debate with such luminaries as Alexander Hamilton and Benjamin Franklin, albeit in chatbot form.

Students in the class take part in two six-week debate-style games on separate teams: patriots versus loyalists and federal power versus states’ rights. They read the works of political philosophers such as Paine, Locke, and Montesquieu, and even dress in colonial outfits. Students take on the role of real characters from history, like James Madison and George Washington, and subsequently create chatbots by feeding AI models with primary and secondary source material read and discussed throughout the course. 

“I read the Federalist Papers and put them into a chatbot for my character, James Madison,” Anabelle Krawitz ’28 said. “It showed me how people can have the chatbot be predisposed to certain types of information based on what we give it, including information about the characters they debated.”

The capstone of the class was pitting the chatbots for Franklin and Hamilton directly against each other. The Hamilton team leaned into demonstrating the power of the federal government. But the Franklin team knew that they could rile Hamilton up, so they fed the chatbot documents about affairs and embezzlement.

“By using technology, we got to the humanity of these two people,” Seidule said. “Students had a blast preparing their respective chatbots, listening to the arguments, and figuring out why they happened. As a historian, I want civil discourse. I want students playing other roles, learning to argue and compete. That’s how we learn social intelligence. The idea of a liberal arts education is not to foolproof students for the technology itself; it’s to prepare them to think clearly, write effectively, work with other humans, and navigate change.”

Carolyn Cotta ’28 agreed. “It was my favorite class I’ve taken at Hamilton,” she said. “People got to interact with the material rather than just studying it. It taught me a lot of skills beyond the classroom. It helped me understand the human element of history. History is very different when you step into the worldview of people who don’t think the same way as you do.”

Asking Great Questions

College Hill has long been a place where diverse academic viewpoints have been expressed and debated, and the topic of generative artificial intelligence proves to be no exception. Around the world, higher educators are grappling with challenges posed by AI tools, like the Large Language Model (LLM) ChatGPT, that can churn out a history essay or physics diagram with just a few sentences of input. But for faculty and students at Hamilton, AI does not need to be a threat to the liberal arts education. In fact, it makes an even better case for that very education.

President Steven Tepper sees a liberal arts education as vital to developing digital fluency and preparing students for this new age.

“We’re entering the fourth industrial revolution, one that is defined by ubiquitous computing,” Tepper said. “But we can’t just use it as a mindless pathway for maximum efficiency. The liberal arts are not about shortcuts; they’re about long takes. We are about purpose, community, and trust. So, how can we use technology at Hamilton to drive more collaborative and connected learning rather than isolated and frictionless learning? Not every college or university is asking that question. Our charge is not just about acquiring and transferring knowledge, it’s about generating purpose and meaning together.”

Tepper describes authentic intelligence as the differentiator that sets humans apart from machines: elements like discernment, judgment, ethics, and aesthetics.

“A liberal arts approach is about asking great questions, seeing what activates the human imagination and spirit — how do we discern in a world of so much information and possibility, how do you prioritize, how do you ask the right question, and not just solve problems but identify problems. That takes acts of human imagination,” he said. “AI can help connect dots faster, but how can we see new constellations and formations of dots? And a liberal arts education builds the ethical foundation that makes sure the human project is intertwined with this technological revolution.”

John Werner ’92, a venture capitalist and senior fellow at the Massachusetts Institute of Technology, sees Hamilton’s superpower as its ability to teach students to see the big picture across boundaries, connecting technology with humanity, ethics, and imagination.

“A lot of the future will be smart people managing fleets of AIs,” Werner said. “Some people will be the bridge from the economy of before to the economy of the future. Superintelligence is around the corner, which will solve most problems in science, medicine, and mathematics. 

“We’re in an era where AI can interact across networks, and software works alongside humans to achieve shared goals. AI will magnify human potential, but also inequality and misinformation if not thoroughly guided. Liberal arts students can play a role harnessing AI for creativity and human values. Hamilton should position students as leaders for the unknown,” he added.

Werner recommends the integration of AI into course curriculums and for students and faculty to leverage tools like Google’s NotebookLM, which can turn a research paper into a podcast that a listener can actively engage with through interactive dialogue. He also noted the importance of team sports to learning, echoing Seidule’s belief that the Hamilton liberal arts education goes beyond the classroom.

“Hamilton is a residential college and students form social groups,” Seidule said. “The classroom is a team. Over the course of the semester, students have four bosses in different classrooms, coaches, parents, and jobs. Living in a dorm, they negotiate with other human beings that they didn’t know before. Residential liberal arts colleges prepare you for life in ways that others don’t. The teaching and coaching staff are engaged with student development.”

The Biological Foundations of AI

Hamilton faculty and staff played a large role in the academic development of Drew Linsley ’08, an assistant research professor of computational neuroscience and artificial intelligence at Brown University. He credits the late Jonathan Vaughan, a professor of psychology at Hamilton, as a key influence on his journey into computing. Vaughan focused on cognitive neuroscience, and Linsley fell in love with the study of biological intelligence, which forms the conceptual basis of the neural networks that power AI tools like ChatGPT.

“The brain is the most complex organ there is, billions of neurons connected through trillions of synapses, and even though there’s been progress in neuroscience, you need to derive some principles of intelligence,” Linsley said. “It happens that artificial intelligence is a great way to do that. A lot of my work is about how we visually reason about the world — what are the algorithms that our brain uses to understand the world, how is that implemented in brains.”

Linsley earned his Ph.D. in 2011, around the time that a breakthrough occurred in computer science in the form of deep learning, a new approach to neural networks, which had existed conceptually since the early days of computing in the 1940s but lacked the amount of data and computing power to be implemented. That changed with the advent of the internet and powerful new computer chips known as graphics processing units.

“What was so exciting for me is that a deep learning model is structured similar to a human brain’s visual systems,” he continued. “It could be trained to recognize images, reason about the world, and with those capabilities, AI transformed from sci-fi to a real and disruptive technology. The second breakthrough was that you could take these models and train them into the chatbots that we rely on daily with a further technique that is similar to the operant conditioning model developed by B.F. Skinner, who attended Hamilton College.”

Linsley notes that many businesses are not ready for artificial intelligence and how it will change their economics and structure. Although he sees younger generations as more fluent in AI than those that came before, he emphasizes that may not be enough to succeed.

“Disruption is not in one business area, it’s potentially all areas at once,” Linsley added. “Going back to my own experiences, Hamilton helped me with preparing for that by being broadly capable so I could follow my own interest. If you asked me why I’m doing what I’m doing, it’s because I wanted to learn how intelligence works. I’ve been obsessed ever since those classes with Professor Vaughan. Students should have a broad skillset and understand where the future is going, then follow what really interests them.”

AI as a Thought Partner

That intellectual pursuit continues beyond College Hill. Zach Kleinbaum ’19, senior machine learning delivery manager at Snorkel AI and previously generative AI strategist at Deloitte Consulting, graduated just a few years before ChatGPT’s release and has seen firsthand the transition as AI becomes more prevalent and valuable in day-to-day business.

“The biggest change is the ready availability of AI,” Kleinbaum said. “I use it as a thought partner. Back in the day, a lot of time would be spent on the hated blank-page problem. Like a paper or presentation, it was hard to start an outline and formulate ideas. AI helps me most with formulating a narrative or idea before I dive in and make it a final product. It helps dump ideas on the page and aggregate them into documents and presentations, as well as automate workflows by connecting all my applications and tools to combine new insights.”

Hamilton faculty are also leveraging AI as a thought partner to improve their teaching. Mireille Koukjian, associate professor of instruction in Arabic, uses AI to help adapt exercises for reading passages and vocabulary drills, and to create lesson plans or PowerPoint presentations for classes.

“This is work that used to take me hours to do,” Koukjian said. “I check it for accuracy, and the human element is still there and not going away, but the speed with which I can do work is different now. I also studied prompt engineering, an important part of AI. Having ChatGPT without knowing how to use it is like having a piano in your living room and not knowing how to play it.”

Learning thoughtful prompting can be a way of teaching students about language. In Koukjian’s Arabic classes, students create stories in Arabic and used AI to generate illustrations. They learn to think carefully about how to describe a scene clearly with the cultural nuances that shape imagery. She emphasizes the importance of setting clear boundaries and acceptable use policies for students in her classes. For example, they are not allowed to use AI to write a paper. She advocates that students have the confidence to question the outputs from AI and recognize their pitfalls.

“I encourage students to read AI-generated material the way they would read any meaningful text to see that cultural nuance can be absent — to see whose voice is being amplified or erased,” Koukjian added. “Let’s say we give AI a translation to do, and it mistranslates an Arabic proverb or poetic line. It’s not a failure; it’s an opening for students that shows that language holds more than information: rhythm, identity, shared memories of a community. Students need to use AI not as a shortcut, but as a collaborator that will enhance their awareness of what languages do. AI becomes a mirror and reflects how humans make meaning and how we still need human creativity to fill pages.”

Anna Huff, assistant professor of digital arts, worked with students to develop a chatbot using the precursor to ChatGPT that behaved like a 4-year-old child, and assessed its creativity.

“I had this question of how an AI learning language might be similar to a child,” Huff said. “I’m a performance artist, not a computer science professor, and I wondered what it would be like to train a model on my own child’s language, then perform the character live. My colleague and I wrote down over 200 lines from our kids, which became our data set. The ways you can push large language models to the poetic and uncanny, that is what interests me as an artist. I ran an AI theatre course in support of the project about how to create a performance algorithmically; similar to an algorithm, theater is a set of rules that we create that lead to performance and creativity.”

Huff emphasizes the importance of the human-in-the-loop. When AI chatbots talk to each other, she sees them spiral into an unrelatable, boring, and inane conversation. Bringing a person into the conversation is what makes it interesting.

“What will become valuable in art is not how many AI buttons you can press, but how you can bring a human back into the loop,” she said. “Humans have a body. I think AI will reinsert the value of the tactile and embody it into the digital art space.”

A Place for Innovation

As part of Hamilton’s commitment to pushing boundaries in research and learning, construction continues on the Innovation Center, scheduled to open in early 2027. The new 41,000-square-foot facility will serve as a campus hub for digital research and interdisciplinary collaboration, providing students and faculty with spaces to work with industry and local partners and integrate technology into every discipline.

“The center is a space but also an idea — a place where all the tracks come together, a laboratory to test new connections,” Tepper said. “It will be a place where regional and global partners can audition innovations for our community to work with in real-time, providing feedback and contributions. I want our faculty to be bold and ambitious with AI-activated learning and discovery.”

Lisa McFall, director of initiatives, scholarship, and collaboration, partners with faculty in developing their research into digital scholarship projects that not only support professors’ research goals, but also create opportunities for students to develop their own digital scholarship methods and skills. She sees faculty and students thinking “outside the box” about ways to use AI.

“They don’t just take the technology at face value,” McFall said. “They find new and exciting ways to use it. Generative AI is a tool, but not the solution for everything. At Hamilton, students develop the ability to have discussions and dialogue to gain a deeper understanding. They bring that critical thinking lens, developed uniquely at liberal arts colleges, where they are encouraged to be independent thinkers.”

Tepper believes that Hamilton can offer a model for the country of a liberal arts education enhanced by technologies like AI and enabled by faculty experimentation.

“We’re in the toddler phase of this technology,” Tepper said. “It’s going to be a teenager in a month. We’re still going to learn a lot and the tools will change. We will give our faculty the flexibility to play; they are the experts in their pedagogy. I would love to see our faculty build the most ambitious liberal arts AI project. That’s my challenge to our community: be bold, ambitious, and clear-eyed and never shy away from useful critique and assessment. I’m sure we will emerge from this with extraordinary insight and momentum.”

Benjamin Hosking is a freelance writer who focuses on higher education and STEM fields, especially engineering and computer science. Outside of news, he consults with college presidents and nonprofit leaders on speeches, books, and major grants.

Totto Renna (a.k.a. Supertotto) is a one-man pixel art design studio based in Napoli, Italy, and Berlin, Germany. Artist, illustrator, and motion and graphic designer, he collaborates with magazines, international advertising agencies, production and publishing houses, and new media agencies.