︎   ︎   ︎   ︎  


Synthetic Media - AI Concepts

Explore the use of Runway ML or another synthetic tool to develop an experiment in storytelling.

︎1. Home Alone (Cat version) - What if your favorite animal talks?︎

Character - Animals / Pets
  • Camera angle - Objective Point View
  • Voice:  AI model works on animal face. 

Model :https://app.runwayml.com/workspaces/?modelId=First-Order-Motion-Model / Deepfake

The basic idea would be to make an embedded animal-face-talking filter for the pet camera so that when a pet owner is not home, they can use the pet camera to communicate with their pets more effectively and with fun.

The story will be a cat version of Home Alone. The main character is a needy, shy, spoiled cat who always wants to be around people. There will be no villains like in the actual movie. One day when she’s alone at home, and something catches her eyes in the drawer. Then she gets stuck in the drawer and cries until her butlers return to the house.

The external feature of a character is a cat (Tom's). Her personality is needy and shy, but at the same time, she wants to be around people. 

︎2. Help Me to Memorize the Components!︎

Character - An ITP/IMA student is struggling to memorize the names of the components for the P comp class. The user will use the camera app that can scan the components and the ai generator will track the confusing features of the components. After the scanning, the names will pop up on the screen next to each component.

Model: https://app.runwayml.com/models/genekogan/DenseCap

Tom and I were talking about how the components from the Physical Comp class are hard to memorize and classify, and we ended up with the idea of captioning the components names with a phone camera. 

︎3.  Hazy / Gloomy photos (landscape) to beautiful landscapes︎

Model: MidJourney / Stable Diffusion / StyleTransfer

Character -  a person who lives in a dystopian world in the future year 2080 or further and want to explore the old world using the ai picture generator.

WHO data show that almost all of the global population (99%) breathe air that exceeds WHO guideline limits and contains high levels of pollutants, with low- and middle-income countries suffering from the highest exposures.

“Air quality is closely linked to the earth’s climate and ecosystems globally. Many of the drivers of air pollution (i.e. combustion of fossil fuels) are also sources of greenhouse gas emissions. Policies to reduce air pollution, therefore, offer a win-win strategy for both climate and health, lowering the burden of disease attributable to air pollution, as well as contributing to the near- and long-term mitigation of climate change.” (https://www.who.int/health-topics/air-pollution#tab=tab_1)

The last idea is about air pollution in the future. People in the year 2080 will not be able to see and feel the clean and fresh air in the story set we built. By using Mid Journey, the AI will generate the landscape of the polluted sky to improved and clean atmosphere and sky.

Response to the Reading Materials

The Supply of Disinformation Will Soon Be Infinite by Renée DiResta

Deepfakes Are Becoming the Hot New Corporate Training Tool by Tom Simonite

The AI Art At Christie’s Is Not What You Think by Jason Bailey

These three articles focus on both the bright and dark sides of AI / Deepfake / Machine Learning. They are questioning whether we can trust AI and whether AI works voluntarily like people.

The article by Jason Bailey also points out that AI makes creativity or the humans who used and generated it as a tool to create art. I think that even though there is a variety of artificial intelligence in the world, but no artificial intelligence makes something from 0. People must provide data or information for AI to learn from the given data. There are mixed evaluations of the artistry of works created by artificial intelligence. Just as humans learn and inspire by seeing, hearing, and experiencing from an early age, AI proves creativity that is no different from human creativity. However, it isn't easy to agree with the idea that the creativity and originality of AI can only be recognized if they meet the standards of human artists.Therefore, any art by AI and people are collaborated work by both, and it will be the audience who values between AI and an artist. People were skeptical about photogrphs when it came out first. But now people are identify photographs and paintings or drawings as different categories in the art.

Moreover, indeed, deepfake is sometimes considered as a tool or portals for prohibited uses. In the article by Tom Simonite, WPP seems to want to change that reputation and fix the bias in data. I also agree that the more people explore uses for AI synthesize imagery, they can make better digitally made models "indistinguishable from real people."