AI and Hyper-personalised Learning
Alright, Since everyone’s out there talking about AI right now, let me share a bit of fun we had with it.
When we started building Web3Compass (Our Web3 on-boarding platform), one of the first things that we did was to integrate an AI-based feedback system for quizzes and assignments in our Web3 courses. We wanted the feedback to be insightful, constructive, and personalised based on user responses.
And, of course, we also added a bit of “personality” to it…
Now, we have an absolutely savage feedback system that’s out there assessing your answers, serving up feedback, reality checks, and mean burns — all rolled into one. [Check the image below for reference]
But beyond the feedback (and the occasional roasts), I think AI can actually help us achieve something far more transformative: Hyper-personalised Learning
To start with, A base level personalisation of content would mean localising it based on language. With the current AI capabilities, it’s super easy to translate and localise content into any given language, making it accessible to a wider audience.
The next level of personalisation can be 𝗰𝗼𝗻𝘁𝗲𝘅𝘁-𝗯𝗮𝘀𝗲𝗱 𝗽𝗿𝗼𝗴𝗿𝗲𝘀𝘀 𝘁𝗿𝗮𝗰𝗸𝗶𝗻𝗴. Unlike usual progress tracking that observes patterns in how you interact with the content, AI takes it a step further by linking your learning pace and challenges to specific topics or sections of the course, giving more context to your behaviour. For example, imagine you’re taking a course on blockchain. Typically, one might just track how many videos you’ve watched or quizzes you’ve completed, but with AI, we can granularize it. We can identify topics or concepts that are easy for you (you might finish those faster) and ones that are a bit hard. Maybe you fly through the lessons on Bitcoin’s history, but struggle with understanding how proof-of-work works. Recognising this as a challenge area, the system can step in and provide personalised support — suggesting additional practice exercises, curated readings, etc. I mean , by that time, it would not be just progress tracking, it would be more on the lines of measuring improvement
Next, Imagine 𝗮𝗱𝗮𝗽𝘁𝗶𝘃𝗲 𝗰𝘂𝗿𝗿𝗶𝗰𝘂𝗹𝘂𝗺 that adjust to your skill level, preferences, and goals. Content that evolves as you learn, offering deeper challenges when you’re ahead of the curve or more foundational guidance when you’re struggling. Even the tone of the lessons can shift — whether you prefer a formal approach or something lighter and more conversational.
A bit more personalisation would mean 𝗼𝗻-𝗱𝗲𝗺𝗮𝗻𝗱 𝗰𝗼𝗻𝘁𝗲𝗻𝘁 that is tailored to your needs. AI could easily create diagrams, videos, or animations to explain concepts in a way that’s easier for you understand. If you prefer hands-on learning, it could provide personalised coding challenges or simulations to practice. For those who like structured guidance, it could generate step-by-step roadmaps based on your progress. Basically generate content that is tailored for you, on the fly.
I guess, with AI and hyper-personalisation , the idea would be to transform content-based learning from a static, one-size-fits-all model into a dynamic, evolving dialogue between the user and the content
Pretty interesting when you think about it !
