One problem I felt very engaged in was the tangible CA project. I enjoyed exploring how a set of simple rules could produce more complex patterns through repeated interactions. It was very mathy, which is what I like. It was very fin working with numbers and different cofigurations, noticing and realizing how some very small and seemingly not important changes could affect the outcome (though I probably should predict because in the end of the day it's still coding). It was the first project, where I didn't really have much experience, in general and with AI, so I learnt a lot through trial and error. A bit frustrating, but very worth it!
However, when moving on to Project 2, it felt somewhat "weird" after the project. I still have not found the sweet spot for the amount of AI I sould use, where I feel like I did not use enough in Project 1, but maybe too much in Project 2. It was fun coming up with the rules, understanding generic algorithm, but once I came up with a solid idea, I realized that if I prompt very carefully, AI can code it up in less than an hour. All the input is what's the expectations about the user interaction, visual presentation, how genetic algorithm works, and the code is ready.Â
In the end, when reflecting on both projects, especially after finishing Project 3, I realized that I feel most engaged when I can experiment, observe immediate results, and make changes based on what I discover. Both problems involved breaking a larger challenge into smaller rules and testing how those rules interacted. The main difference was how I came to the final product. With the tangible CA project, I spent more time working through problems myself, making tons of mistakes. With the genetic music project, AI made it so fast to move from an idea to a n implementation, but that also meant I had fewer opportunities to discover things through the coding process itself. I don't think using AI is necessarily a bad thing, but these projects made me think more carefully about when it helps me learn and when it takes away part of the experience. I want to find a balance where I can use AI to get "unstuck" or explore possibilities without giving up the experimentation and problem-solving that make a project satisfying to me. It also makes me wonder that when and if one day the AI bubble does burst, what would be the future for engineers?