Deep Learning: What's behind Donald Trump's Twitterbot?

Posted by John Onwuegbu | March 06, 2016 |

Twitter handle @DeepDrumpf, created by Brad Hayes, a postdoctoral associate and AI Researcher at MIT, uses deep learning inspired by an existing training model, as well as a recent report that analyzed the presidential candidates’ linguistic patterns to construct Donald Trump speeches at a fourth-grade level.

While deep-learning is a field of artificial intelligence (AI) that uses neural networks to teach computer to to find patterns on its own.

Trump's Twitterbot creates Tweets one letter at a time, essentially learning through an underlying structure from all the data it gets, and then comes up with different combinations of the data that reflect the structure.

The bot, for instance, may randomly begin to create Tweet with the letter M, somewhat likely to be followed by an A, then K, and so forth, until it types out Trump’s campaign slogan, "Make America Great Again." It then starts over for the next sentence and repeats the process until it reaches the 140-character limit.

Albeit, Trump’s Twitterbot doesn’t always make complete sense with its Tweets, but at least its partially coherent - much like the candidate himself.

It interface with Trump’s Twitter account, allowing the algorithm to tap the language from the real Trump’s Tweet, which primes it to give a response that is even more likely to be contextually relevant. 


Brad, however, thought that this would be a good way to learn more about some of the concepts of AI, and have a little bit of fun in the process.