Inspiration
College reading assignments can sometimes be devastating, especially when there is a large block of code with nothing but words on it. Some of the reading assignments can be 5000-10,000 words a week, which makes it almost impossible to get started. My biggest fear about reading a big article like this is mainly two things.
- I don't want to read it because it's too hard and too boring.
- I read it, but after reading it, I still know nothing.
So to solve this problem. We created this brand new software called ICANTREAD to better help people read long articles with less pain and more gain.
What it does
The app mainly have two functions targeted two different concerns:
- Better and faster reading experience:
We create both text-to-speech translation and image generation to give more information through visual and audio systems, so readers can better grasp the idea of the passage.
- Better understanding of the reading.
We create a summary title at the beginning of each paragraph and a general summary at the beginning of the article to help the reader understand the logistics of the article. We also created an interactive quiz at the end of the article so that the reader can test his understanding based on the quiz result.
How we built it
Challenges we ran into
Figuring out how to make the generated quiz better – we needed the quiz to be in a specific format in
Order to parse it correctly.
Using python & flask together, a new tech stack
4.running out of AI generation credits
Accomplishments that we're proud of
The app works and is completed!
We think that we will actually use and refine this in the future – look out for the official deployment
What we learned
Sometimes vanillaJS is pretty convenient for making quick prototypes
It’s easy to lose time getting bogged down in specific tech problems related to the platform, but if one solution doesn’t work, we can try to think of 3 more and hopefully one will solve the issue
How to call apis
ChatGPT can code really fast
What's next for ICANTREAD
Better control output of AI-generated images – how can we ensure that they are accurate?
Simplify / change the tone of each paragraph
Light and and paper dark mode
Log in or sign up for Devpost to join the conversation.