Research · Project
AI Tools for the Workers Algorithms Manage
AI in the workplace is usually built for the firm. This NSF project asks what it looks like when it is built for the workforce.
Online freelancing, in which professionals work on a collection of individual tasks outside traditional workplaces, is rapidly becoming a significant component of modern labor markets. Freelancers need to market themselves, find well-paying and skill-building work, and perform and get credit for quality work. The design of online labor markets often makes this hard: their policies tend to favor employers, and their algorithms for matching people with jobs and prices are opaque to workers.
The idea
AI-enhanced digital tools may be able to help workers manage their profiles, workload, and task performance. The question is what those tools should do, and who they should answer to. Rather than optimizing for the platform, this project starts from the needs, values, and constraints of freelancers themselves.
How we work
The project uses human-centered and value-sensitive design. Working with a carefully selected and steadily updated sample of online freelancers, we gather data through interviews and focus groups to identify the functionality workers need, then design, deploy, and refine prototypes over multiple cycles. Throughout the three-year project we collaborate with organizations dedicated to supporting online workers.
Coverage of the award: Skidmore researchers seek to democratize AI tools to support online freelancers.