Scaling the Internship

Engineering Social Media for Good

UC Santa Cruz Assistant Professor David T. Lee’s Tech4Good Lab at the Baskin School of Engineering’s mission is to modernize the way we teach, the way we work, and the way we solve problems as a society by combining social networking technologies and small group dynamics to create little pockets of intimacy and productivity.

“Social media is almost anti-social,” Lee said. “Interactions are so brief. You toss out a comment and leave, meaning there are no face-to-face interactions, meaning there are no social norms that you’d get in an in-person dynamic. So one of the things we’re interested in is whether or not it’s possible to create a network environment where you can have discussions about controversial things like abortion or the environment without everyone yelling at each other.”

Common values usually exist between groups of people but are often difficult to detect on social media. Why not make social media more interpersonal and relational?

As a researcher studying Electrical Engineering in the Social Algorithms laboratory at Stanford University, Lee studied mass decision making and created mathematical models describing the polarizing effects of social media. As he mastered the tools of social computing, Lee started asking whether he could use them to translate the empathy and intimacy of small groups and scale it up.

Lee’s group is studying methods of bringing small groups of people together and letting them share goals and values together before they discuss controversial issues. These small group deliberations are then networked together to seed a larger-scale conversation.

It’s a radical new way of looking at conversation online. But political and ethical deliberation isn’t the only social nexus the Tech4Good Lab plans to improve and scale-up: They’ve just completed a pilot program for a project called Causeway that takes on-the-job training and translates it into bite-sized online chunks called micro-roles.

The technology offers a potential solution to some of the trickiest problems in the workforce: How you obtain opportunities to learn when you don’t have the experience to get an internship in the first place? How do you add value (and earn a living) as you learn the fundamentals of a new skill?


Causeway teaches its users computer science by dividing the subject matter into smaller, project-based roles. For example, the pilot program taught web development. As a user learns to perform a certain task, they are assigned a small piece of an existing project. The Tech4Good lab has proven that novice web developers using the platform could learn and retain web development skills while creating a number of static websites for non-profit organizations.

“There are tons of online courses and they’re all superb at delivering one aspect of the educational experience: content,” said Lee. “Causeway addresses the other aspects of learning.”

Traditional Massive Online Open Courses (MOOCs) such as Khan Academy, Coursera, Udemy, and EdX provide hundreds of thousands, even millions of users access to education, but they’re focused on providing content, which is important but incomplete.

“Think of the experience you get working on a real project during an internship or in a lab,” said Lee. “We’re trying to imagine what a MOOC for an apprenticeship or an internship might look like. Where if I had a couple of hours in the evening or on the weekend I could pick up a micro internship and get some real world learning that sequences towards learning a professional skill.”

Internships and other types of apprenticeship learning (such as on-the job training) take time and resources, not just from the student, but from the mentor and the institution they’re working for. Causeway would allow people to start adding value to existing projects right away while improving their skills.

Read his the paper…

Causeway uses a concept called micro-role hierarchies. The platform breaks the complex task of creating a web site down into discrete chunks (micro-roles), while tracking individual progress to ensure that each student is proceeding toward his or her learning goal.

“Causeway breaks down specific projects so learners can work on real world projects even if they don’t have a background in computer science or web development while also establishing a hierarchy that allows users to move up (learning new skills and mentoring others below them by answering questions or reviewing code),” said Veronica Rivera, a Ph.D. student in the Tech4Good Lab.

Learners log onto the Causeway platform, read guides and after picking up a few basic skills will be able to do a single useful function. These tasks are divided up and distributed across groups of learners.

“During the pilot program we got people with no knowledge of web design to build static pages for nonprofits,” Rivera said. “Our next step is adding chat, code reviews and other ways of making it more user friendly.”

There’s no strict sequence to the learning, which means a swarm of students tackle websites simultaneously, aggregating skills, assembling dozens of websites piece by piece.

Career Progression in the Online Gig Economy

Causeway has attracted national attention. In 2019 Lee was invited to the White House to provide his perspective on apprenticeship learning for the JobKit Developers Conference. There, academics and representatives from the private sector, education and government all met to discuss how to prepare the American workforce for the future.

Automation and artificial intelligence may consume a growing fraction of traditional employment possibilities. People are unlikely to be obsolesced, despite scare stories to the contrary. What is more likely to happen is that more and more employment (particularly entry-level) will involve checking the work of and filling in the gaps left by artificial intelligences and robots. Acquiring new skills and hands-on experience will of course be evermore important for these workers.

Education isn’t the only issue with ad hoc labor (i.e., “the gig economy”) however. A huge problem that the Tech4Good Lab is also trying to address is career progression in online labor markets.

One of the projects Rivera is working on looks at ways of making career growth possible in the gig economy. “Besides studying how we can scale computer science education (by using Causeway, for example), I also study ways we can enable career progression in online job markets like Amazon’s Mechanical Turk,” she said.

Online labor markets will continue to grow as automation spreads through different industries. Indeed much of the work being done involves filling in the gaps left by artificial intelligence: labeling images, transcribing data. But as more people enter these labor markets, they’re working on such small, disparate tasks that they can never really progress in their careers.

“So we started looking at career paths in online labor markets and comparing them to traditional jobs,” Rivera said. “We’re finding that a big part of career mobility is balancing the tension between earning and learning, and leveraging social structures that exist in the context of complex work. Since social structures don’t exist in online labor markets, workers can only really make choices and tradeoffs between earning and learning. But the many structural problems in online work make it next to impossible to balance these.”

The Tech4Good Lab is hoping to design platforms that will enable complex work and social structures for workers in the online gig economy. That ethos extends to the Tech4Good Lab.

“I joined the Tech4Good Lab because I really wanted to do something that made a difference,” Rivera said. “And I think education and career growth are important topics, especially for the future. But another reason why I joined the lab was the collaborative atmosphere–we aren’t just doing data research, we’re really making a difference and we have dozens of undergrads in the lab helping us out.”

Professor Lee has made it his mission to accept all students who want to do research in the Tech4Good Lab. During the Spring quarter, 20 students helped out with their projects, a massive boon for undergraduates for whom research opportunities (which often involve carefully cultivating a relationship with a particular professor and developing a deep interest in a very narrow subject matter) can be sparse.

In other words, the structure of Lee’s lab is a social experiment designed to address one of the biggest problems undergraduate researchers face. This move toward constant improvement for the greater good is encoded into the lab’s name.

“The reason why I called it the Tech4Good Lab was that I never wanted to lose focus on what I was doing,” Professor Lee said. “It’s too easy to get stuck on a single approach or locked into a single academic community. My lab is defined not by its method or approach but by its impact.”

It’s a new kind of lab and a new approach to societal problems. Are you interested in helping out? For more information please visit:

Leave a Reply

Your email address will not be published. Required fields are marked *