Holi, Soy Anita. 

I’ve spent most of this series talking about frameworks, checklists for choosing AI tools, why-how-what filters, questions to ask before you adopt anything new. All useful. None of it is the actual reason I’m doing any of this. So let me back up and tell you the story first, and let the Why show up where it actually belongs. At the end, earned, instead of stated upfront like a mission statement nobody asked for. 

Where it started 

I taught Mandarin. Not in one tidy classroom for a decade across different systems, different kinds of students, from young kids to adults studying for HSK exams. Every context asked something different of me. What worked for a classroom of curious eight-year-olds was not quite associated with an adult learner trying to pass a proficiency exam on a deadline. I had to keep re-learning what “teaching well” actually meant, because the answer kept changing depending on who was in front of me. 

Somewhere in there, I also spent time evaluating AI model outputs, doing unglamorous, detail-obsessed work of checking whether an AI’s answer was actually good, or just confident-sounding. Prompt engineering and edge-case testing, before either of those had the job titles they have now. 

And more recently, I moved into Product Marketing and Partnerships in Tech building things from scratch such as a vendor pipeline, an internal operations system, a way of tracking a product from validation through development. Different industry entirely. Same underlying instinct, it turns out. 

The question that kept following me around 

Here’s what I didn’t notice until I sat down to actually name it: every single one of those chapters was quietly asking the same question. Not “what tool should I use” or “what’s the trend.” Something underneath that. 

How do you design an experience that actually changes what someone can do afterward?

That’s the thread. Teaching Mandarin, evaluating whether an AI’s response actually served the person asking, building a vendor pipeline or an internal system from scratch. All of it come down to the same design problem, just wearing a different outfit each time. Did this actually change what the person on the other end is capable of or did I just look like progress? 

Which brings me to my Why 

I use technology and learning design to make education more meaningful, accessible, and future-ready. 

That’s not a slogan I picked because it sounded good on an application. It’s the sentence that finally explained, after the fact, why I kept ending up in rooms that looked nothing alike but kept handing me the same question to solve. 

I think a lot of teachers reading this have their own version of that thread, even if no one’s ever asked them to name it. If you’ve moved between roles, or between countries, or between “teacher’ and something else entirely, and you’ve quietly wondered whether any of it actually connects. It probably does. It’s usually just waiting for someone to ask the right question. 

I’m going to keep unpacking pieces of that thread here and in the newsletter with The Shine Lab including where it’s taking the AI Agent Toolbox I’m building next. If any of this story sounded familiar, that’s who I’m building it for. 

Subscribe to keep thinking this through with me

— Anita