Pressure Project: Audio Visualizer

The goal of our first pressure project was to create a self-generating patch, given the time limit
of five hours. I was stuck at first, due to the openness of the project. But, I knew that when
starting in the class, I wanted to make an audio visualizer, so that is what I tried to do.

The time constraint didn’t feel like a massive issue, because I knew this pressure project was
more about exploring the different tools and seeing what we could make with a base
understanding. I had a very loose understanding of touch design, with only a few basic nodes
under my belt, so I knew I needed to start by watching some tutorials if I wanted to make a full
blown audio visualizer. I watched a few different videos, letting me a better understanding of
how certain audio and visual nodes work. I was able to mess with the values along the way, and
combine certain knowledge from each video to make something that was mine. I could choose
what I felt looked best and where in my composition, giving me so much room to work with.

What surprised me the most about this first pressure project was the sheer amount of different
nodes you can work with in touch designer. Given all of these nodes, and all of the different
combinations of the nodes, I can’t imagine how many different programs you could run with this
application. I felt like I grew a lot and got more understanding of the potential and opportunities
with touch designers, working mostly with the audio nodes and how they can affect the visual
nodes, but I am really looking forward to what I can do with the application in the future and how
far I can push my knowledge of it!

#2 AI Expert Audit: Attack On Titan

Briefing Doc: Here
Character Mind Map: Here
Podcast: Here

For my investigation on NotebookLM, I decided to do an overview of both the Attack on Titan the manga, and the anime. I imputed the sources of both the wikis for the different mediums, and a few other articles on bigger episodes of the series. I chose Attack on Titan because it is a piece of media that I have obsessed over for awhile now, due to its intense world building and focus on deep storytelling between characters, and an overarching story. Now after inputting these sources, I found that AI had a good grasp on what the story is, but isn’t able to know finer details without me letting it learn other sources. I want to go over some of the parts that AI got right, and the parts that AI got wrong.

The AI had a good grasp on what the general idea of the story was, along with the characters and what they represent. But, it had mistakes on certain things like pronunciation, character connections, and small, yet key parts of the story. It oversimplifies a lot of the parts of the story, leaving out what really makes the show intense and intricate. For example, I generated the mind map to create connections between characters, yet it just divided them into sub-group, diminishing what the depth of character connection there is in the story. 

I believe that if I didn’t know much about the story, it would be a good way for me to get a basic understanding of it. I found the podcast to be the most useful, and entertaining way of learning the material because it feels more engaging and goes further in depth. The mind map didn’t really do much, and the briefing doc was informative, but not engaging. The idea of it being AI made me feel weird in itself, especially with how more human AI is sounding, with expressions and banter. I told the AI to make the voices engaging and funny, which made the podcast feel more enjoyable. With these human emotions though, it wasn’t felt with the words they said always, a lot of it sounded scripted, along with occasional voice glitches that took me out of it.

After going through this research, I found that it is a new possible way to learn about more specific topics that there might not be many resources for. Even then, I wouldn’t use it as something to rely on. You have to make sure your sources you’re uploading first are good sources, and even after that, it can focus on the wrong information if you don’t prompt it correctly. If they were able to find a way for it to take up less resources, along with more accuracy, I could find myself using something like this. The medium that led me to feel this way is the podcast option. It was very engaging when it wasn’t glitching out, and it dove deep into the storytelling and world building that I asked it to, and the information was mostly accurate. I think it has the possibility of being a big learning medium compared to things like ChatGPT that just spew information back out at you.