In 2010, as students at the Media Lab, Karthik Dinakar SM ’12, PhD ’17 and Birago Jones SM ’12 collaborated on a class project to create a tool to assist content moderation teams at companies like Twitter (now X) and YouTube. The project garnered significant enthusiasm and led to an invitation to demonstrate the tool at a cyberbullying summit at the White House. However, they encountered challenges in getting the tool to function properly.
After struggling to create a working demo the night before the White House event, Dinakar reached out to Jones to express his frustration. Upon examining the data, Jones discovered that while Dinakar’s model was identifying concerning posts, it was missing nuances in teenage slang and indirect language used by the posters. This realization highlighted the importance of having individuals who understand the data involved in building machine-learning models.
This insight inspired the development of user-friendly tools that allow nonexperts to build machine-learning models, leading to the creation of Pienso. Pienso now enables users to develop language models for detecting misinformation, human trafficking, weapons sales, and more without the need for coding.
Reflecting on their journey, Jones and Dinakar emphasized the significance of empowering domain experts to create models tailored to their unique data sets. This approach has been instrumental in their collaborations with various organizations and institutions, including using Pienso to assist in understanding and combating emerging challenges such as Covid-19.
By providing a platform that allows users to build customized AI models quickly and efficiently, Pienso aims to promote a future where AI solutions are developed by those with the most intimate knowledge of the problems they seek to solve.
