Natalie Lao initially planned to follow in her parents' footsteps as an electrical engineer, until her path took an unexpected turn. Enrolling in course 6.S192 (Making Mobile Apps) taught by Professor Hal Abelson, she discovered a revolutionary approach to technology. This course revealed how smartphones could transform into powerful tools for locating clean water sources, organizing facial recognition systems, or accomplishing virtually any task imaginable. "I realized that if more people understood technology could be this accessible and impactful, it would change everything," Lao reflected during a break from her dissertation writing.
Transitioning her academic focus at MIT to computer science, Lao became part of Abelson's innovative lab, which was actively disseminating its App Inventor platform and empowering do-it-yourself philosophy to high school students globally. This experience with App Inventor ignited Lao's passion for understanding artificial intelligence for beginners and making it accessible to diverse populations, from agricultural workers to factory employees. Now completing the final year of her PhD at MIT, Lao has also co-established an AI startup dedicated to fighting misinformation and serves as co-producer of an educational series focused on machine learning tutorials for non-experts. These initiatives all align with her overarching mission to help individuals discover their inner innovator and critical thinker.
"Natalie embodies an infectious blend of optimism and enthusiasm," notes Abelson, the Class of 1922 Professor in the Department of Electrical Engineering and Computer Science (EECS). "She possesses innate leadership abilities that inspire and organize people effectively."
Lao immersed herself in App Inventor, developing educational modules to teach students how to construct facial recognition models and implement cloud data storage. However, the 2016 U.S. presidential election prompted her to examine technology's role more critically. Rather than political concerns about Trump himself, Lao was more disturbed by evidence suggesting that social media-amplified propaganda and misinformation had significantly influenced the election's outcome.
When colleague Elan Pavlov, then an EECS postdoc, approached Lao with his concept for AI tools for combating fake news, she was immediately ready to contribute. Having experienced life in rural, urban, and suburban areas of Tennessee and Ohio, Lao was accustomed to diverse political perspectives. However, she observed how social platforms were filtering these voices while amplifying polarizing and often inaccurate content. Pavlov's approach particularly resonated with her because it focused on identifying the individuals (and automated accounts) spreading misinformation rather than merely evaluating the content itself.
Lao enlisted two friends, Andrew Tsai and Keertan Kini, to help develop the platform. They eventually named it HINTS, an acronym for Human Interaction News Trustworthiness System, inspired by an early page-ranking algorithm known as HITS.
During a demonstration last autumn, Lao and Tsai showcased a network of Twitter accounts that had disseminated conspiracy theories connected to the murder of Saudi journalist Jamal Khashoggi using the hashtag #khashoggi. When investigating what else these accounts had shared, they discovered streams of other false and misleading content. Among the most prominent was the baseless assertion that then-U.S. Congressman Beto O'Rourke had financially supported a migrant caravan approaching the U.S. border.
The HINTS team anticipates that by identifying networks responsible for spreading fake news, social platforms will accelerate the removal of fraudulent accounts and curb the dissemination of misinformation.
"Fake news remains ineffective in isolation—actual people must consume and share it," Lao explains. "Regardless of political affiliation, we prioritize factual accuracy and democratic integrity. Misinformation campaigns target all ideological perspectives, exacerbating political divisions."
The HINTS team is now collaborating with their inaugural client, a Virginia-based media analytics company. As CEO, Lao has leveraged her project management experience from internships at GE, Google, and Apple, where she most recently oversaw the implementation of the iPhone XR display screen. "I've encountered few individuals as adept at managing both personnel and technology as Natalie," remarks Tsai, an EECS master's student who first met Lao while serving as a lab assistant for Abelson's course 6.S198 (Deep Learning Practicum) and now serves as HINTS's CTO.
As HINTS was gaining momentum, Lao co-founded a second venture, ML Tidbits, alongside EECS graduate student Harini Suresh. While learning to construct AI models, both women became frustrated with existing YouTube tutorials. "They were dominated by complex equations with minimal visual aids," Lao notes. "Even when the concepts aren't particularly difficult, the presentation makes them seem intimidating!"
Confident they could create superior accessible AI education platforms, Lao and Suresh reimagined intimidating topics like unsupervised learning and model-fitting as approachable "side dishes." Sitting cross-legged at a table in a fireside-chat setting, Lao and Suresh put viewers at ease with real-world examples, whimsical illustrations, and an engaging presentation style. Six additional videos, funded by MIT Sandbox and the MIT-IBM Watson AI Lab, are scheduled for release this spring.
If her audience takes away one message from ML Tidbits, Lao hopes it's the realization that anyone can grasp AI's fundamental principles. "I want them to recognize that this technology isn't exclusively the domain of professional computer scientists or mathematicians," she emphasizes. "They can learn it too. They can develop informed perspectives and participate in conversations about its appropriate application and regulation."
