Getting a Head Start on AI

The Davis Institute for Artificial Intelligence and Halloran Lab for Entrepreneurship host SureStart, a summer program that teaches real-world skills to students interested in AI

Nafis Bhuiyan at his computer
Nafis Bhuiyan '26 reviews his pitch before presenting during the SureStart Makeathon event in Ostrove Auditorium. Bhuiyan worked as a summer academic research assistant at Colby's Davis Institute for Artificial Intelligence
Share
By Bob KeyesPhotography by Michael Seamans
August 21, 2023

A new generation of ethical AI designers spent eight weeks at Colby this summer for an artificial intelligence boot camp.

The 13 students from Colby and partner school Lincoln University in Pennsylvania learned essential AI skills, then put those skills to real-world use by developing entrepreneurial business models and pitching their ideas to the community.

Administered by Colby’s Davis Institute for Artificial Intelligence and the Halloran Lab for Entrepreneurship, the summer-long SureStart curriculum at Colby evolved from a longtime friendship between Amanda Stent, director of the Davis Institute for Artificial Intelligence, and Taniya Mishra, founder and CEO of SureStart.

SureStart is a national model for building pipelines of diverse tech workers through AI skill training and project-based learning. Established in 2021, the Davis Institute for Artificial Intelligence is the first cross-disciplinary institute for artificial intelligence at a liberal arts college.

“Taniya has built a wonderful curriculum, and it has been my pleasure to bring that curriculum to Colby,” Stent said as she introduced the students and their projects during the SureStart Makeathon Aug. 4 at Ostrove Auditorium in the Diamond Building. 

One of Colby’s newest specialty labs, the Halloran Lab for Entrepreneurship provides students with opportunities and experiences to innovate and explore the process of creating solutions to problems they identify. Jeremy Barron ’00, the lab’s inaugural director, said the SureStart program provided the new lab with an ideal project because it involved students solving problems with new technologies that are full of opportunities for practical innovation.

“It’s been an amazing collaboration between SureStart, Davis AI, and the Halloran Lab, and I hope this will be the first of many partnerships with Amanda and with others on campus,” Barron said.

Getting started in AI

In most undergraduate programs across the country, the first opportunity students have for taking a course in artificial intelligence is their junior year, Stent said. “And that is a shame because at that point you have eliminated a lot of potential AI tool users and tool builders. We think AI belongs to everybody, and SureStart helps ensure more people have those opportunities.”

In addition to nine rising sophomores from Colby, the program attracted four students from Lincoln, the nation’s first degree-granting historically black college or university. Lincoln’s president, Brenda A. Allen, is a member of Colby’s Board of Trustees.

In the first phase of the program, the students spent five weeks learning hands-on AI skills and becoming familiar with AI coding tools like TensorFlow and Hugging Face. With the support of mentors, each student developed individual ideas to pitch.

Based on voting on the individual pitches, the students broke into four cohorts, refined their ideas, and settled on one pitch per cohort. At the event on Aug. 4, each cohort presented a well-researched, market-designed idea for using artificial intelligence to solve a real-world problem with convenient, cost-effective solutions.

The ideas included app-based products to help people manage their migraine headaches by tracking and analyzing data; to make it easier for students to search efficiently for college funding or scholarships; to create a more convenient way to shop for clothing online to reduce returns; and to create generative 3D models that could be useful for architects, designers, or hobbyists.

A panel of judges and mentors evaluated the ideas and offered feedback and suggestions.

The pitches

Miz Insigne ’26 was part of a cohort that included Desmond Frimpong ’26 and Peggy Jones ’26. They pitched a web-based application called ScholarSheets, which would use ChatGPT, an AI natural language model, to help students simplify the search for colleges and scholarships by making it possible for them to filter and sort based on their profile and eligibility.

“I’ve been interested in artificial intelligence for a long time, but it’s hard to figure out where to start if you don’t have guidance. When this opportunity came up—I heard about it from a professor—I said, ‘That sounds amazing.’ I am grateful I got the opportunity to learn about all of these things.”

Miz Insigne ’26

To make the case for the ScholarSheets, Insigne posted an image of the spreadsheet they created during their college-search process—a dizzying jumble of multi-colored columns filled with data, words, and symbols indicating the status of various applications.

“That slide is my actual, real college-application spreadsheet document that I worked on in 2020 and 2021. It has over 70 columns of every single scholarship and college that I applied to,” Insigne said during the pitch. “This single document took me months of work. Of the 60 schools I applied to, I ended up with two realistic choices. All of that time and effort is not something every student can afford, especially those who have similar backgrounds as me—an international and low-income student.”

Audience members applaud after listening to Miz Insigne ’26, Desmond Frimpong ’26, and Peggy Jones ’26 pitch a web-based application called ScholarSheets, which would use ChatGPT, an AI natural language model, to help students simplify the search for colleges and scholarships.

Though most of the SureStart students are computer science majors, Insigne is double majoring in science, technology, and society and in mathematical sciences. They said SureStart appealed to them because of their growing interest in artificial intelligence.

“I’ve been interested in artificial intelligence for a long time, but it’s hard to figure out where to start if you don’t have guidance,” they said. “When this opportunity came up—I heard about it from a professor—I said, ‘That sounds amazing.’ I am grateful I got the opportunity to learn about all of these things.”

A cohort of David Roberts ’26, Tait Kline ’26, Sir Noble Appiah-Dankwah ’26, and Michael Adekunle, a Lincoln University student, pitched their idea called MyGraine, an app that would collect data about sleep quality, diet, local weather, and other factors to help predict the onset of migraine headaches among individual users.

“I have personally experienced migraines for the majority of my life, and throughout that time I have struggled to know why they happened to me,” Kline said during the pitch, adding that “powerful prescription” medications never produced lasting relief. “Once a migraine started, there was almost no way to escape the pain, and the uncertainty of it terrified me. Now that we are beginning our journey into AI, the time is right to help other people in the same situation.”

Cathy Fan ’26, Michael A. Yilma ’26, and Nafis Saadiq Bhuiyan ’26 presented an idea called Ptereo, a subscription-based service to produce 3D models. Using generative AI, users would input text into a prompt, and the program would convert that text into a set of images and finally a 3D model suitable for virtual reality or 3D printing. The team’s market research revealed many complaints about current 3D-modeling programs, including the cost and speed, as well as steep learning curves.

“We want to streamline this 3D-design process and involve anyone who is able to bring their own ideas to 3D,” said Fan.

A trio of students from Lincoln University—Jannah Burkes, Ravyn Pleasants, and Ite Ibitoye—pitched the idea My Perfect Fit, which would utilize virtual models to help individuals order clothing online that fits well and matches their expectations while maintaining their privacy. Their research showed that clothing retailers lose $63 billion annually because of online clothing returns.

An AI-powered tool, My Perfect Fit would help make online shopping more efficient for consumers and retailers alike, said Burkes. Perhaps most important, it would ease the frustrations of shoppers, especially those who live in rural areas and have little choice but to shop online to find the clothes they love and want.

“We are very familiar with the struggles and frustrations involved in the online shopping practice,” Burkes said. “There have been many times we have been excited for our clothes to arrive, and they are not what we expect.”

Pleasants, a rising junior at Lincoln, said the SureStart program at Colby exceeded her expectations. “I wanted to learn more about AI, and I did that. But I also got a lot of knowledge, connections, and resources,” she said. “I got more than I could expect, honestly. I feel like I will use everything I learned this summer at Colby.”

Constructive criticism

It remains to be seen if the ideas generated by the SureStart cohorts will ultimately receive funding and have a chance to prove themselves in the marketplace. But the initial feedback was helpful in finding their way forward to the next step in the process.

“We think AI belongs to everybody, and SureStart helps ensure more people have those opportunities.”

Amanda Stent, Director of the Davis Institute for Artificial Intelligence

SureStart Makeathon mentor Matt Hawkins ’17, an engineer at Lockheed Martin, returned to campus to watch the presentations. He said he was impressed, and he expressed hope that the Halloran Lab will help the students and their ideas maintain momentum as they move into the next phase of development, including facing hard questions about their viability.

The real world is a harsh place for new companies, said Hawkins, who, in addition to his work in the aerospace industry, is building a Boston-based startup in the real estate sector with other Colby alumni.

“It’s great to see the energy and enthusiasm of the students, and they did a great job,” he said. “But I am going to be critical in my feedback. I don’t learn much from ‘Nice job.’ I learn from criticism, framed in a helpful way.”

Other SureStart Makeathon mentors were Lisa Kaplan ’13 from startup Alethea, Josh Kim ’23 from investment organization MaineAngels, and Geoffrey Poon, data scientist at Bloomberg Second Measure. Stent and Barron said they were looking forward to involving more Colby alumni and partners with student-directed work on campus, building the extended Colby community around AI and innovation.

related

Highlights