Week 9: Challenges and Reflections
Avaya A -
Hello everyone!
After eight weeks of working on Meta Chameleon, from setting up models and datasets to making predictions and drawing connections to the real world, it feels surreal to say that I’m nearing the final stretch. This week, I wanted to take a breath, reflect on the entire process, and share some of the challenges I’ve faced and what I’ve learned along the way. There were plenty of bumps in the road, but also moments that reminded me exactly why I started this in the first place.
Setting up Meta Chameleon wasn’t as simple as I’d initially expected. I ran into problems almost immediately, such as with the Python package’s compatibility, the confusing documentation, and models that wouldn’t load or train correctly. There were moments where I questioned whether I had taken on something too complicated. But I learned to break big problems into smaller ones, troubleshoot through forums and GitHub threads, and ask for help when I needed it.
I also had to work hard to understand the “why” behind the code. It’s one thing to get an AI model to make predictions, but it’s another to understand why it’s making them. Early on, I focused too much on getting results, and not enough on thinking critically about the input data, model behavior, or bias. It wasn’t until Week 3, when I really dove into hypothesis-building and research design, that I started to appreciate the science behind the model and why a strong foundation matters just as much as fancy predictions.
Once Meta Chameleon started giving outputs, I realized another challenge: What do I do with this information? How do I know if it’s accurate, useful, or even interesting? This pushed me to start comparing my results with real-world outcomes, researching how professional analysts measure model performance, and even thinking beyond sports (like in Week 8, where I explored how AI can save lives). I started to see my project not just as a tool, but as a conversation starter – a way to think deeply about how we use AI to understand the future.
I’ve learned that perseverance pays off, and that understanding your data, your goals, and your model structure is more important than just getting fast results. AI is powerful but imperfect. Predicting outcomes is never completely accurate, and there are ethical and technical limitations I’ve become much more aware of. Reflection is key. Writing these weekly blog posts has helped me organize my thoughts, track my progress, and stay grounded in what I’m building.
As I get ready for my final post next week, I’m thinking about how this project has challenged me, changed the way I think about AI, and opened up possibilities I never expected when I started with just a blog title and an idea.
Comments:
All viewpoints are welcome but profane, threatening, disrespectful, or harassing comments will not be tolerated and are subject to moderation up to, and including, full deletion.