🧠 Modules 2 & 3 Complete: ADA Applications, Perspectives, and Problem Fit

I just completed Modules 2 and 3 of Vanderbilt University's ChatGPT Advanced Data Analysis course on Coursera—both with full marks.

These modules focused on:

• Real-world ADA use cases: working with small documents, structured data, and media

• Automation with .zip files: using compressed files to scale repetitive tasks

• Turning conversations into tools: reframing interactions into executable workflows

• Evaluating problem fit: understanding what types of tasks ADA handles best

While I can’t say I encountered anything radically new or unique, that’s not a complaint. Having worked closely with ChatGPT and ADA for nearly two years, many of these techniques were already familiar. What’s helpful—and genuinely interesting—is seeing alternative angles on problems I’ve already tackled. Sometimes a change in framing opens the door to new efficiencies.

One takeaway that’s been growing clearer for me is this:

The most valuable knowledge isn’t about what ADA can do, but about understanding how LLMs work—and what they can’t do.

Knowing the limits of language models is what allows meaningful, efficient, and realistic problem design. If you understand how these systems generalize, interpret, and infer, then you stop trying to force-fit them into roles they’re not suited for—and start applying them where they shine.

Among the few new things I did learn:

• ADA can accept and work with .zip files, enabling batch workflows. That’s a game changer I simply hadn’t run into before.

• The image manipulation capabilities within ADA are more advanced than I expected—definitely worth exploring more.

So while these modules weren’t revelatory, they were solid. I’m continuing with a sense of curiosity, and looking forward to seeing whether Modules 4 and 5 dive into more complex or unexpected use cases.