Research
Position papers, studies, and systems work. Everything here is published to be examined, reproduced, and built upon — including the parts that did not go the way we expected.
AI is evaluated by whether it completes tasks. We argue for a complementary standard: whether it expands the time and capacity people have for judgment, learning, care, and shared work.
Amara Osei, Daniel Reyes, June Park
Most AI interfaces assume one person and one model. Most meaningful work happens in groups. Design patterns from a year of building shared-context systems for small teams.
June Park, Tomás Carvalho
A benchmark tells you what the model can do. It tells you nothing about what its users can still do. First results from paired capability–capacity evaluations.
Daniel Reyes, Priya Raghavan
Learning tools concentrate where money and English already are. Notes on building systems that adapt to learners across languages, bandwidth, and prior schooling — by design rather than retrofit.
Noor Haddad, Amara Osei
Research speed is constrained less by ideas than by trust in one's own results. The unglamorous systems we built so that every experiment is reproducible by default.
Lena Vogt, Yusuf Demir
We instrumented the working weeks of 140 knowledge workers. Less than half their time went to the work itself; the rest went to finding, translating, reconciling, and reporting. A map of the overhead.
Tomás Carvalho, Priya Raghavan