For this assignment, I designed a propaganda poster to push the agenda of my course character Hwei.
I went with louder design language, finding an image of a robot colored in bright red and oranges, and then overlaid that with text highlighted with red and pink outlines. The white text seemed like a good choice to contrast from the red and orange robot and to make it more legible.
As for the context, it is a message being sent to all of humanity to embrace the coming of AI and the robots being produced by Hwei. He believes that these robots are the future of society. Rather than assimilating the robots into human culture, Hwei believes that robots are a higher being altogether, and should replace the current reign of humans in the “biological” food chain. Although this would most likely get rid of the biological part of civilization entirely, there would be no pain, death, or suffering. Every robot would be programmed for the greater good of the earth, regardless if this new world has any room for humans or not. Accept it, don’t fear it!
As Hwei ventured on his quest for world domination through the harnessing of AI powers and creating life through his research, Aggressive Technologies was a corporation that he never saw eye to eye with.
Aggressive Technologies looked to take the AI market by storm, engulfing all of the industries pertaining to, e-commerce, social media, among other fields of tech. Their sole focus on expansion and obsession with creating a monopoly has made their corporation one that seeks to trample upon others to gain the upper hand.
On one hand, Hwei may seem to have the same overall goals as Aggressive Technologies, taking over the world and having the power to shape anything he touches. However, it has never a goal of his to gain monetarily from his research and creations. In fact, his ideal new world would consist of no monetary value on any item, idea, or being. All he ever wanted was to create the way he wanted to; for the world to see the potential beauties of the world from his perspective. The purpose of his new AI beings would not be for his gain, but rather for the betterment his world (at least for his perfect creations).
To be fair, some of Hwei’s equipment and parts of his laboratory were stolen from Aggressive Technologies, as he sent out his rogue thief programmed robots to infiltrate their headquarters. The masterminds behind Aggressive Technologies know about the research of Hwei and the existence of some of his true to life robots, but have not identified him as a serious threat yet. There will soon be a clash between the two parties, as Aggressive Technologies is threatened by Hwei’s unwillingness to back down to their expansion, while Hwei fights back against the irrational development of AI in his eyes, ruining the beauty of his craft.
For my choice of assignment this week, I decided to do the Talk to the Bot assignment. I actually did the Consult with your doctor as one of my optional assignments for a previous week, so I already had experience generating speech from Dr. Oblivion and creating a narrative. I decided to make my course character, Hwei, ask Dr. Oblivion to become one of his evil sentient robots with a body that would aid him in taking over the world by creating perfect AI beings that would take over humans.
Learning how to generate the lip synced video was a whole different process compared to the original Consult with your doctor assignment. It wasn’t hard, but it really did take a while and I think having a higher end PC and good internet speeds expedited the process for me. The process was streamlined enough to where I only had to click a few buttons and create the right folders. The tutorial given was very helpful for me to follow directions and create the videos.
What really intrigued me was the Python code used for the entire process. There was a “show code” button for each of the processes, and I looked through lines while I waited for my videos to generate. I was fascinated to see how each piece of code worked for the application. and it was definitely some good future personal project ideas for me in the future.
For the Dr. Oblivion clips generated, I was a little disappointed at how the lip sync wasn’t all that great. I’ve seen some other examples at the top of DS106 weekly assignment posts, and those look to be synced a little better than mine. There were parts where at the end of a period, the Doc’s lips kept moving and it felt a little jarring to me. For Hwei’s voice, I used a basic speech generator online to make mp3’s of his voice. I tried to use an AI voice replicator to take the original League of Legends game voice for hwei, but many of them required a paid subscription and was out of reach for me. I spliced it all together by using the Clipchamp online video editor, and it only took a few minutes to put the video, audio, and Hwei’s avatar picture together.
Overall, I had fun with this assignment and I learned a lot about AI generation for audio and video. Hopefully I can even make an AI bot myself with my own code! I’ll make sure to cater it towards taking over the world with AI, or else Hwei might hold my life against me
This is hands down my favorite ever project for DS106.
Initially I had a hard time coming up with a film to analyze. I don’t have any subscriptions to any streaming platforms, so I had to do my best to find clips on YouTube. While brainstorming, I had a lightbulb moment that reminded me of a particular scene in WALL-E that I wanted to delve further into.
I picked a scene where the majority of the attention is put on the characters. I focused in on the close-ups of the characters, and how different cinematography elements were strung together to tell a story without the use of dialogue.
I used OBS to recording my desktop screen, with the audio recording on. I wrote a script beforehand and printed it out to read, and it was recorded all in a continuous take. This allowed me to not perform any extra work of overlaying my voice on the clip, nor did I have to download the video to my computer. However, this forced me to record everything in a single take, which gave me no leeway to mess up any of my lines for 3 minutes. It took way too many takes than I want to admit to get all of my lines down, and timing my commentary with the transitions of the clips was even more difficult.
Once I did get it down though, I was so satisfied with the end result. I think this is one of my best works so far, and I had a blast walking through the creative process behind the analysis. I’d argue that the timing of the commentary is a method of storytelling in and of itself, and I think it really shines through in my WALL-E analysis.
I’ll definitely explore editing clips in a video editor in the future, but learning OBS was a great experience for the rest of my DS106 journey.
The Middlebury college AI assignment was an interesting read and activity. As an introduction of the assignment, there was an in-depth discussion about the regulation of AI, and how the development of free and accessible online tools has created issues for original creators and artists. The widespread recognition of Stable Diffusion is something I’ve come across in my time on the internet as well. It’s becoming increasingly difficult to differentiate the difference between real art and AI generated ones, especially for already animated content. If I were an artist (which I am most definitely not), I wouldn’t be happy with the direction of the industry and I would frankly be worried about how this changes my creative direction for the future. I’d assume that many artists are stuck at a crossroads between pursuing a type of art they’re truly excited about versus adhering to the AI buzz and losing some originality in the process.
For the actual image generation portion of the assignment, I decided to create an image of a dachshund on the moon, and I formatted the prompt to where the dog would resemble an astronaut, akin to Neil Armstrong during the first moon landing. I decided to use both Stable Diffusion and Craiyon to generate these images.
This first one from Stable Diffusion was amazing, 10/10 in my book. The image feels very realistic yet still with an artistic flair. The generator even stayed true to the prompt and added an American flag on the left arm sleeve. The only criticism I’d have would be the weird lighting coming from what I’d assume is the sun? Doesn’t seem very realistic for there to be beaming sun on the moon. Otherwise, it kind of scares me how good this image was.
This one from the Craiyon generator is something that I expected a little more. The image is very low quality and doesn’t adhere to the guidelines I set in the prompt. It’s almost comically bad, the moon completely overtakes the subject of the photo which was supposed to be the dog.
I didn’t get any submission result back after sending in my results on the website, but I will update here if there is anything else that pops up from this adventure.
