“A breakthrough in machine learning would be worth ten Microsofts.”
- Bill Gates
HUMIC is concerned with fostering and promoting machine learning education, projects, and research on the Harvard campus. We provide opportunities for students to learn about machine intelligence in a community environment and promote interdepartmental and multi-university interaction relevant to machine intelligence. Our club brings those interested in machine intelligence closer together, and to explore the possibility of applying machine intelligence tools to multiple disciplines.
HUMIC has two backbone offerings:
The first is the Future of Intelligence Fellowship (FIF). FIF is a semester-long deep learning workshop series that serves as an accessible on-ramp for members to learn machine learning concepts
in an interactive and social setting. We offer a beginner track for those new to machine learning and an advanced track for those with existing deep learning knowledge and who seek to extend their knowledge past the fundamentals. Students learn about everything from neural networks to CNNs to transformers to diffusion models and reinforcement learning, and will get involved in meaningful projects from day 1 that they can showcase on their GitHub. We believe teaching is the ultimate form of learning, and so all of our workshops are student-created and student-led. An essential component of the fellowship is the built-in mentorship program. Affectionately deemed "mini-batches", these are small groups of students led by a senior member who serve as an initial nuclear community that welcomes and catapults students into the study of machine learning. They go out for various fun events, and become an essential resource for any questions related to machine learning and life in general. Furthermore, we invite Harvard's leading professors as well as industry professionals to give talks and treat our fellowship members with private dinners to form valuable professional connections that will last a lifetime.
The second is the AI Project Incubator. HUMIC sponsors a set of semester-long machine learning projects that demonstrate innovation and creativity at its finest. Selected students gain unparalleled support and mentorship in building real-world AI solutions. Different projects require different degrees of experience in AI/ML. Past projects include a course recommender chatbot, contributions to DeepMind's open source libraries, a Sushi Go Party! reinforcement learning agent, AI-generated comic books, and more! supports student projects that apply machine learning in any capacity.
These are student-initiated, student-led, and student-driven machine learning experiences that benefit directly from the wealth of knowledge provided by the HUMIC community and network.
In addition, we regularly hold speakers events with luminaries in AI: industry leaders, AI entrepreneurs, and research experts. We have had the honor of hosting Daniel Gross (founder of AI at Y-Combinator),
Surya Bhupatiraju (research engineer at DeepMind), Susan Murphy (a MacArthur Fellow/"Genius Grant" and professor of RL here at Harvard), and others. Typically, a lottery is run to get a private dinner opportunity with these trailblazers.
Some of our less periodic programming includes graduate-level discourse for advanced undergraduates. The Theoretical Research Team gathers a few highly motivated students for an ambitious target: published research in a top machine learning conference like NeurIPS, ICML, and AAAI. This team gains intimate knowledge about the very frontier of machine learning and together challenges the community's understanding of learning in novel and innovative ways. They represent the best of Harvard's minds working together to push the boundary of machine learning scholarship. The ML Reading Group is a community of students who assist each other in their learning goals. They meet every week to discuss a selected paper, both classic and cutting-edge. By discussing a paper in a supportive and knowledgeable community, students come out with a deeper understanding of the content, and are often more excited about the material.
Most importantly -- HUMIC is a COMMUNITY. Learning AI/ML is hard, but the good news is that there are so many talented people in HUMIC to help you along. We prioritize making this a space for you to find invaluable friends and upperclassmen mentors, in AI and in life in general. Want to push yourself to read papers? Join a reading group or make an accountability friend. So what are you waiting for? Join us now! HUMIC warmly welcomes you to join us in our endeavor to uncover the secrets behind machine intelligence.
Special Events Director