The idea of internet censorship and the possibilities of modern natural-language processing to an authoritarian agent.

What it does

Monitors text chat on a Discord server and uses Google's sentiment analysis API on Google Cloud to track statements relating to a number of nouns. Upon detecting a user expressing positive opinions of "banned words", the bot will generate propaganda posters using an AI and send them to the user, and also activate the project's hardware component - a spooky face which will open its eyes and shine a light from them.

How I built it

The Discord bot and Cloud components are built using Node.js, with the hardware control and Propaganda AI using Python.

Challenges I ran into

Audio processing proved to be too difficult, necessitating a change to text processing.

Accomplishments that I'm proud of

The construction of the robotic face is probably the neatest hardware design our team has ever come up with, and the Propaganda AI is also a very interesting component.

What I learned

A frightening amount about audio formats.

What's next for Famous Last Words

Analysing some of the weirder qualities of the propaganda slogan-generating functions.

Built With

Share this project: