Project Ideas – Blake Almon
For the sophomore year the project idea for i2 which I lean towards the most this year is fully automated youtube channel that is controlled by a trained machine learning algorithm. I hope to allow the algorithm full access of tools, information, and the youtube channel itself to see what it would do to achieve both viewers and subscribers. This project would help shine a light on the potential of machine learning while possibly helping content creators create videos with the ease of a click.
Another idea I had which would be most likely a backup idea is “white paper hub,” a website where one could compare and contrast white papers while having a seamless feedback giver similar to Grammarly when writing white papers. This website would also hold an archive of white papers while allowing one to scan them for similar ones and find ideas that no one has through a machine learning algorithm uses them through retrieval augmented generation. This project would help students, teachers, and other people who may be writing or using white papers/research papers.
How I came up with these ideas?
-After learning some machine learning basics, I was opened to the possibilities of AI and what it could do. From this new found view of the world I was able to combine some of my interests with machine learning to achieve new ideas that would possibly be used for my 2023-2024 i2 project.
What do you think your biggest challenge will be in the early stage?
-I think my biggest challenge for either project would be the machine learning itself. To be more clear, for the automated youtube channel, the hardest early part of the project would be figuring out a model and setting up the embeddings to make sure the model works the way I want it to. Then for white paper hub, I believe the earliest hardest challenge will be figuring out a way to score each paper using retrieval augmented generation in real time when someone edits a paper.