The Berkeley Operations and Behavioral Analytics Lab runs real-world experiments with workers and companies to understand how people use AI at work, and builds tools to help them work better.The Berkeley Operations and Behavioral Analytics Lab runs real-world experiments to understand how people use AI at work.
Algorithms assign work, give advice and increasingly act on their own. Using field data and experiments, we study how people learn alongside these systems and how to design them better.
Algorithmic Advice
How algorithms should guide multi-step decisions so the advice helps now and still leaves people better on their own.
When should advice name the exact next move, and when should it teach a broader rule?
Exact advice works best while people can see it. A broader rule keeps working after it is gone.From Precise or Broad?
AI at Work
How professionals actually use generative AI day to day, what it changes about their own work, and where it still falls short.
How do people use generative AI at work, and how much should AI agents decide alone?
Generative AI speeds up familiar tasks and makes new ones worth trying, though switching costs hold some teachers back.From The Teacher's Knapsack
Learning on the Job
How workers build skill alongside an app's recommendations, and how platforms can hand out work so people get better.
As people learn a job, do they lean on an algorithm's recommendations or outgrow them?
Newcomers follow the app's job recommendations. Seasoned couriers increasingly go their own way.From Learning on the Go
Platform Work
When and where gig workers choose to log on, and how platforms can pay, motivate and keep enough of them working.
What brings people to work through a platform, what keeps them there, and how should they be paid?
Pay matters, but so do habits: many drivers stop soon after reaching an earnings goal.From The Impact of Behavioral and Economic Drivers on Gig Economy Workers
Setting prices, scheduling shifts, deciding when an AI agent should hand a customer to a person. We design the study with your team, run it in your own operations and hand back measured results and tools you keep.
Example projects
The kinds of questions companies bring us, from forecasting demand to when an AI agent should hand off to a person.
Orders delivered on time
+13.6 pts
Example result
We test the change at random in your own operations, so you see what it did, day by day.
AmazonRouting, delivery and warehouse work, where algorithms plan the day
Research that makes algorithms easier to work with
Our goal is to make the algorithms people work with easier to learn from and to trust. We started with gig workers taking jobs from apps, and now study AI agents and the assistants people rely on every day.
A research grant from the Fisher Center for Business Analytics at Berkeley Haas
Grant, Jul 2023A talk on the future of skills in the AI era, at NASA Ames
Talk, Sep 2023Two papers on human-AI decision‑making, presented at MIT
Talk, Nov 2023Two papers on human-AI decision‑making, presented at Stanford
Talk, Mar 2024For Park's operations course in the Evening MBA program
Award, May 2024How K-12 teachers use ChatGPT on their own work
Accepted to CSCW, Sep 2024Runner-up for the INFORMS TIMES Best Working Paper award
Award, Oct 2024Two papers on how people and algorithms make decisions in sequence
Talk, Nov 2024Managing Multihoming Workers, selected for the 2025 symposium in Boston
Talk, May 2025Reinforcement learning that turns decision data into tips people can use
Published in Management Science, May 2025How gig workers learn alongside recommendation algorithms
Accepted to CSCW, Jul 2025Learning on the Go and Making ChatGPT Work for Me, presented in Bergen
Talks, Oct 2025How advisors respond when their clients also consult a personal AI
Preprint on arXiv, Mar 2026For teaching in the Evening & Weekend MBA program
Award, Apr 2026Whether algorithmic advice should name the next step or a broader rule
Major revision at M&SOM, Apr 2026Dispatching jobs to gig workers who learn and can turn jobs down
Major revision at M&SOM, May 2026Three papers presented at Harvard Business School
Talks, Jul 2026Dispatch policies that build gig workers' skills
Posted on SSRN, Aug 2026Where an AI assistant's memory of its user goes wrong
Preprint on arXiv, Oct 2026How firms can learn about pricing without running costly live experiments
In progressCompliance-aware reinforcement learning for managing human agents
In progressHow to staff fluid teams in a fragmented labor market
In progressDoes skill learned in one job carry over to another?
In progressHow to design and sequence incentives for platform workers
In progressHow K-12 teachers use ChatGPT, and how it changes their productivity and the tasks they choose
In progressLooking for a specific paper? You'll find all of them on our publications page.
All publicationsPublications
Recent papers from our lab
SSRN working paper 4924786, 2026. Frames lesson preparation as picking the quickest set of tasks that meets a learning goal. A year-long case study of 24 teachers suggests generative AI helps both by speeding familiar tasks and by making it worthwhile to switch to new ones, though switching costs hold some teachers back.
Proceedings of the ACM on Human-Computer Interaction, 9(7), 1–35, 2025. Draws on more than a million delivery orders to show couriers get faster and more punctual with experience. Newcomers tend to follow the app's job recommendations, while seasoned workers increasingly go their own way.
Major revision, SSRN working paper 5387320, 2025. Asks how a platform should offer jobs when workers may decline and improve at different, unknown rates. Proposes a tractable benchmark and learning algorithms, including a variant that caps the risk of market failure, and tests them on delivery-platform data.
The people of BOBALAB
Our alumni have gone on to graduate programs and firms like these
Alumni
