AI startup develops out of Chinese internet regulations

While the rise of artificial intelligence might be a scary prospect for anyone who has seen a Terminator movie, businesses in the D.C. area are taking advantage of the technology’s possibilities.

Local startup Thresher focuses on analyzing large volumes of data, and CEO Becky Fair says the use of AI and machine learning came from the study of Chinese internet censorship.

When trying to create a model of China’s system of regulation, researchers realized they were missing a large part of the conversation.

“The problem is, the Chinese citizens themselves were using code words to hide from the Chinese government,” said Fair. These code words, then, were missing from the analysis.

“Finding those words at scale of hundreds of thousands, or millions, of social media posts was really hard. So [researchers] created this computer-assisted approach to surfacing these words, and we turned that into a tool that we can apply to all sorts of language problems,” Fair told What’s Working in Washington.

Thresher applies this approach to sets of language data that aren’t what traditional programs are trained on.

“A lot of the tech today is doing a great job at a lot of the standard language: think the Googles of the world, training all of their models against Wikipedia,” said Fair. While that approach works for pages such as Wikipedia, it doesn’t handle neologisms, memes, slang or jargon very well.

“This is so valuable because understanding and communication is really hard, it’s something that we as humans struggle with, let alone getting computers involved,” she said.

Listen to entire June 19 show:

Copyright © 2024 Federal News Network. All rights reserved. This website is not intended for users located within the European Economic Area.