Harvard’s New Science Dean Wants Students to Tackle ‘Genuinely Hard Problems’

Harvard College Dean of Science Jeff W. Lichtman wants to reshape how undergraduates are solving problems.

Lichtman — who was appointed as Dean of Science in July — said in an interview with The Crimson that Harvard must rethink how the sciences are taught in light of generative artificial intelligence models like ChatGPT.

“There’s a whole bunch of interesting things happening now in the world that never happened before that really challenge the fundamental paradigm in the way we teach science,” Lichtman said. “Almost any question you can ask in science that is already known the answer to, you can get if you have a cell phone or you have a laptop.”

To adapt to these new circumstances, Lichtman said the sciences division should focus on teaching students to pursue “genuinely hard questions” – questions that can’t be answered by the internet or AI.

“Learning to cope with things we don’t know, as opposed to trying to get the answer to a problem set question where the answer is known, is a completely different world, and that’s the world that science lives in,” he said.

Lichtman said the division is discussing the potential for a program that guides students towards answering research questions even before they start taking classes freshman year. The program will pair students with a professor or a tutor who can provide insight on what classes they should take or research they can pursue to help them answer that question.

“In math and statistics, there are genuinely hard problems, in even astronomy, every one of our departments has genuinely hard problems,” Lichtman said. “Having students find a genuinely hard problem that interests them, they know nothing about it, and then have a mentor, perhaps a professor as a tutor who guides you on what types of courses you might think about taking and then for the next four years try to figure out the answer to that problem.”

In addition, Lichtman said the program has potential to be credit-bearing to ease the academic workload and make it more accessible to students with busy schedules. He said he hopes to launch the pilot program next year.

Most likely students won’t be able to answer their questions, according to Lichtman. Instead, they’ll walk away with a better understanding of science.

“Now, will you be successful? If you are, you get a Nobel Prize,” he said. “The chance is you won't be successful, but at least you’ll understand what it is to be in the sciences.”

Answering genuinely hard questions can also influence students to think about the real-life applications of required courses that otherwise feel abstract, according to Lichtman.

“It is a bit of a challenge to learn things when the purpose is to learn it well enough to get a good grade, but that’s it — that’s the only reason you’re learning it,” Lichtman said. “As opposed to just jumping through the hoops, you want to be motivated about why you’re jumping through these hoops.”

Lichtman said that teaching college students to think more about unanswered questions also has benefits for the broader world of science.

“I think young people have a real advantage if you turn education on its head and ask them to deal with problems that are hard,” Lichtman said.

“In fact, not knowing anything actually gives you an advantage because you can think outside the dogma that has not worked to figure out these problems,” he added.

—Staff writer Caitlyn C. Kukulowicz can be reached at caitlyn.kukulowicz@thecrimson.com.

—Staff writer Andrew Yu can be reached at andrew.yu@thecrimson.com.