In 2010, Karthik Dinakar SM ’12, PhD ’17, and Birago Jones SM ’12, students at the Media Lab, embarked on a project aimed at helping content moderation teams at major companies like Twitter (now X) and YouTube. Their work ignited a great deal of interest and they were invited to present their project at a cyberbullying summit at the White House, but the tool had to be completed first.
Dinakar worked hard to devise a working demo that could flag worrying posts on Twitter. But unfortunately, the model wasn’t accurately detecting the content due to the usage of youth slang and other indirect language. This led to the realization that the model builders should be the ones who best understand their data, and not just machine-learning engineers.
This valuable insight guided Dinakar and Jones to develop point-and-click tools that enable non-experts to construct machine-learning models. These tools evolved into Pienso, a platform that is currently helping users to build extensive language models for identifying misinformation, detecting human trafficking, weapons sales, and more, without needing to write code.
Jones and Dinakar started their work on Pienso as part of a Natural Language Processing course and continued till they completed their master’s degrees in 2012. They were twice invited to the White House to give a demonstration of their work, but they did not focus full time on Pienso till Dinakar completed his PhD in 2016, when deep learning started gaining traction.
The two founders credit MIT’s Industrial Liaison Program (ILP) and Startup Accelerator (STEX) for helping them find early partners like SkyUK, which used Pienso to create models to comprehend their customer’s main issues. Later, when the Covid-19 pandemic began, government officials sought Pienso to enhance their grasp of the emerging disease, the platform assisted experts in virology and infectious disease in setting up machine-learning models to scrutinize numerous coronavirus research articles.
Pienso operates on internal servers and cloud infrastructure, and provides an alternative to businesses which may be pressured to share their data with services offered by other AI companies. Earlier in 2021, Pienso announced a partnership with GraphCore, a company that provides a faster, more efficient computing platform for machine learning. The founders assert that this partnership will further lower the hurdles to leveraging AI by considerably reducing latency.
The goal of Pienso is to facilitate a future where more efficient AI models are created for particular use cases by the people who know and understand the problems they are trying to solve best.
