From Safety to AI: Making Tech Click

Picture this, about four years ago, I was the safety gal. I wrote and delivered those mandatory training sessions, you know, the ones everyone has to sit through. This was “for” employees at tech and biotech companies in California. Endless compliance stuff, the same material year after year. I tried my best, I really did. I'd polish the visuals, break down the regulations into plain english, anything to keep people awake. Most of it was still pretty dry. I'd see those glazed over expressions and just think, There has to be a better way.

Fast forward to now. I'm a UX designer working with developer tools and AI, and I'm also taking a learning design course in a UX graduate program at ASU. I feel as if all the pieces are starting to connect. I can see my path straight from those mind numbing compliance sessions to the intuitive, user friendly experiences I'm building today. The intersection of where good design meets effective learning gets me excited. That's where I see myself growing.

I've always been drawn to helping people learn. It doesn't matter if it's through a 1:1 tutoring session, summarizing a 30 page policy into a one page infographic, or hosting a lunch and learn. Right now, that passion plays out in my work with the Generative AI Collective in San Francisco. I help put together events for founders, engineers, and anyone curious about AI. I run panel discussions, hands on workshops, that kind of thing. 

My path into learning design didn't start with textbooks or lectures. It was more of a trial by fire. I learned by tackling real-world problems. For example, how to make an injury report form less confusing, a software tool easier to use, or an AI system less scary. A few years ago, I worked as founding designer at Codemod, a startup building an AI-powered tool to modernize old code. It was a huge learning curve. We were using machine learning to help migrate codebases. I was responsible for running usability tests, figuring out where developers were getting stuck, and designing solutions to make the process smoother. Good design isn't just about making it look nice, it is also about teaching users how to use a powerful tool like Codemod without needing a massive instruction manual.

Since then, I've been diving deep into the world of Generative AI tools like Midjourney, Replit, Clay, HeyGen, Bolt.ai, Apollo… the list goes on. Some of them are incredibly well designed, and others, well, not so much. What really struck me was how often the onboarding and learning experience felt like an afterthought. It was as if these companies were saying, "Hey, check out our amazing technology! Good luck figuring it out!" I'd spend hours playing around with these tools, making mental notes about what worked and what didn't. 

And that's what I'm passionate about now: the place where UX, AI, and learning all come together. I want to help people not just use these powerful tools, but learn to master them all while enjoying the process in between. I believe there's so much untapped potential in creating better first impressions, documentation that doesn't put people to sleep, and learning systems that adapt to the individual user. Why can't an AI tool guide you step-by-step, understanding your needs and adjusting its approach, instead of just throwing you into a chatbot with little to no context? That's the challenge I want to tackle.

I think my unusual background gives me a unique perspective. I've been the sole injury responder for 37,000 employees at Meta, dealing with everything from minor cuts to major emergencies, always needing to provide clear, calm instructions. I've designed products in the fast paced, often chaotic world of startups, built internal tools for biotech companies, and created platforms for immigration services. Through all of this, I've developed a knack for simplifying complex information. And that, I believe, is the foundation of both good UX and effective learning design – cutting through the noise so people can actually focus on what matters.

In the short term, I'm really interested in working on projects that make technical tools more accessible. Long term, I want to play a role in shaping how we teach people how to get value out of Generative AI, creating onboarding experiences that are engaging, not overwhelming, tutorials that adjust to your learning style, and systems that adapt to your individual needs. I imagine a future where using AI feels intuitive and empowering, not like some complicated thing only tinkerers can truly unlock the value of. I am hoping this learning design course will help me refine my thinking and deliver real lasting value. 

One specific area I'm really excited to explore is adaptive learning, powered by AI. Remember the safety training I used to give? They often fell flat because they were one-size-fits-all – rigid and easily forgotten. But what if we could change that? Imagine an AI tool that recognizes you're a visual learner and automatically shows you a diagram instead of a block of text, or slows down the pace if it senses you're feeling overwhelmed. I've seen glimpses of this in tools like Duolingo, which adjusts its lessons based on your mistakes. I want to bring that kind of personalized learning to the AI tools that are transforming the workplace, combining technology with a human touch to make learning truly effective.

I'm inspired by people like Julie Dirksen, author of "Design for How People Learn." In particular, I find it insightful to keep sight of the concept of a behavioral gap rather than a knowledge gap. Great learning design can go beyond delivering content and step into the world of driving behavior change in action in the real world. Also, feedback and early wins are very important, which is important to incorporate in onboarding and throughout foreverboarding. For example, Duolingo seems to do a great job of understanding where the users are at today, giving them personalized exercises and keeping users on track with goals and badges moving forward. For example, I’ve heard my sister multiple times want to stay on track and not “lose her Duolingo Streak” 

Acknowledgements: I would like to acknowledge the use of AI-assisted writing tools (Gemini, ChatGPT, and Grok) in the process of composing this blog post. These tools were used for brainstorming, refining phrasing, expanding on ideas, and checking for grammatical correctness. The core ideas, experiences, and opinions expressed in this post are my own.