BullTwit!

BullTwit logo
Our team!
Example of our real tweet VS AI generated tweet
Example of our real tweet VS AI generatied tweet

Inspiration

We wanted to take the negativity that is fake media, and use it for a positive to bring people together. Also, Twitter is sometimes a parody of itself so this is funny (kinda?) and entertaining.

What it does

Players are given a tweet, which can be real or generated by an AI. They have to guess whether the tweet was real or not.

How we built it

We used react js for the front end and flask for the back end with python. We used socket io for connecting the ends together. We used GPT-2 for generating the fake tweets based off of the real tweets. We have a file of real tweets, which we query and generated based off of it.

Challenges we ran into

We had a lot of dependencies, hence it was difficult to manage and deploy. It was difficult to keep track of where data was going, hence integrating all the separate parts was difficult.

Accomplishments that we're proud of

Generating the fake tweets, the UI, and the logo!

What we learned

How to use sockets to link front end and back end.

What's next for BullTwit!

Make selections have a countdown, for example, when the Twit Master chooses the fake tweet, they have a certain amount of time to choose the tweet. Otherwise if the timer runs out, their points are deducted.
Dark mode!
Have user profiles so users can see their scores from previous sessions
Have a chat function where players (excluding the Twit Master) can converse to each other when making decisions on which tweet they think is fake