Yesterday I read Grace Carney’s blog post on what USV has learned about AI learning over the past year, and it got me thinking about two ways I’ve been using AI tools to speed up my own self-directed learning through adaptive learning lenses and bespoke learning environments.
Adaptive learning lenses
When I was 13, my dad taught me all the countries and capitals in Africa through a fun game of Jeopardy! with my stuffed animals as co-hosts. This process took time, patience, and plenty of human effort. Today, with AI, we can instantly create custom learning tools that make information relatable and easier to remember.
If I had to learn the same thing now, I’d probably design a Taylor Swift-inspired GPT-bot to match country names with her lyrics. Here's how I’d start:
By bringing a relatable personal element (ie: my stuffed animals) into the story, I adapted my learning lens to help me more easily find patterns and associations in the material that resonated with me.
But learning today extends far beyond the classroom. Today, I use AI to help me on nearly every aspect of my self-guided learning journeys. I've curated my ChatGPT instance with a range of hyper-personalized micro-apps tailored to specific learning goals. These include a writing coach, a mock interview simulator, mentorship from public figures I admire, and domain-specific chats that help me troubleshoot business challenges. I adapt my lens based on what I'm trying to learn.
I’d love to see easier ways to generate custom-curated, character-driven, or thematically inspired learning lenses that help people (both current students and also mid-career adults like me) learn best. By the way, this variety is also what keeps learning fun, and dynamic, no matter how old you are.
Bespoke learning environments
Back in school, I’d start the year by buying fresh pencils, notebooks, and binders, getting my learning environment set up for the year ahead. Now, when I start a new job, I don’t buy notebooks—I build a custom GPT to organize and process the information I need. As I add more data and documents, the chatbot becomes richer and more helpful, which deepens my understanding and also helps me be more effective in the role.
Today, AI tools serve as the modern equivalent for these school supplies—empty at first but full of potential to capture learning. Like setting up a developer environment (which might include a code editor, compiler, and debugger), learners can customize their own AI-powered learning spaces. A learning environment might include a custom GPT for summarizing documents, a persona-driven chatbot for answering questions, visualization tools for complex concepts, and a real-world challenge generator to apply learning outside the classroom in real time.
This could also radically change the paradigm for how teachers work with students to facilitate differentiated learning pathways. Imagine if teachers collaborated with students at the start of the school year to help students design AI-powered personalized learning assistants, rather than simply offer a deluge of homework assignments and due dates. These micro-apps could range from role-model-based tutors (e.g., “learn math from Taylor Swift’s ticket pricing strategy”) to “shell” AIs with empty brains that grow smarter as students feed them information—just like filling a notebook with ideas.
Rather than being tested on memorization, students could be evaluated on how effectively they've trained their AI, how well they've organized the information, the creativity they've applied, and the accuracy of the AI's outputs. In the working world, I can easily imagine a future where candidates proudly submit custom GPTs or AI bots to hiring managers, offering a more dynamic and project-based alternative to the traditional resume or portfolio. (Honestly, I've considered doing it myself.)
Embracing this technology—rather than shying away from it—could also bring a more playful, engaging element to the process of learning something new. After all, in a way, the AI is training us, too. We are learning how to think more creatively. Or, as Taylor might say, we're "ready for it."