The week of Designing with Extended Intelligences was extremely interesting and eye opening for me. Designing with AI is something that I have always found daunting and have only played with it at a surface level. Until this seminar, I had not used any AI except the occasional ChatGPT and DALL-E.
The activity with the DOTTOD Camera was a really intriguing way to learn about prompt writing. Not being able to see the original image was a good way for the prompt ideas to be less constrained, specific and limited to what we think we can generate from the image.
I also liked the ModMatrix project as it gave me clarity on the process that the intelligence uses and the tools to set up similar synthetic generators with more specificity in it's purpose
Before this week, my use of generative AI was limited to explorative generations and text-to-image generators for when I have an idea of what I wanted the output to look like.
It was interesting to identify the underlying biases in different APIs. I got more of an understanding of what the latent space is, how the model works in a predictive and hierarchical manner. This helps me have more clarity on how to write prompts and how to train models int he future.
What I learned this week was the main focal point of the project we did for the Micro-challenge 2. I will definitely be using everything I learned in my future here at MDEF and further in my life thanks to Pietro and Chris. Thanks for making sure every big chunk of scary information was digestible and thanks for answering all my questions with patience and making me feel comfortable enough with the concept to freely explore.
AI is a huge part of my design space now. I would like to work with it to generate speculative generations, decision-making, to make abstract and artistic representations of boring and complicated data. I really enjoyed this seminar. I look forward to learning more about this new and evolving technology.