How to Personalize with AI

You might call it hyper-personalization, adaptive learning, specialized scaffolding or differentiated curriculum. In reality, these are all terms to describe the process of creating unique, engaging learning experiences to meet the learners where they are, which is one of the core principles of learning design. AI has emerged as one of the leading tools to spearhead new technologies for chatbots, AI tutors and other similar tools which enable personalized learning experiences.?

With the right data sets, AI models can be trained to leverage user data in designing differentiated learning materials based on learner strengths and weaknesses. One such successful case study is Geniebook - its AI designs hyper-personalized materials for students based on data insights gleaned from their online learning platform.?

The following are the three core areas which I think learning professionals should focus on in order to personalize experiences with generative AI.

#1 Content

The ability to personalize content for different learner profiles and needs is what I like to call the "many-to-many" concept, a term I'm borrowing off the marketing sphere. Essentially, "many-to-many" means using AI to generate multiple personalized marketing messages. In summary, “one-to-many” refers to using the same marketing messaging being sent to a large consumer audience. As you probably have inferred, then “many-to-many” means to develop differentiated messaging for the many different members of the consumer base, with the messaging deeply personalized based on insights curated through data analytics.?

Similarly, curriculum designers and educators can leverage gen AI to generate multiple versions of the same content, whether tweaking the content format or the difficulty level, to meet the needs of the learners. The intention is to drive positive affect and boost motivation - too easy and the learner gets bored; too difficult and you turn learners away. Personalizing the content helps you target each learner’s zone of proximal development in order to stretch and engage them sufficiently.

#2 Attention

Disadvantaged learners may not have access to the personalized attention or time from teachers / educators which more privileged learners may have. Gen AI can bridge this income and privilege gap - users are able to access real-time learning support, as long as they have access to the Internet. Stuck on a concept? Access your AI tutor immediately for guidance in real-time, to guide you through your specific questions through targeted coaching.?

In addition, the AI tutor can personalize its response, explanation and questioning to the user's input, depending on how well the user has displayed his / her grasp of the concepts being taught. An advanced learner may have more scaffolded

#3 Feedback and Assessment

Feedback is a core part of learning - and also a part of the experience which is riddled with time constraints.?In an online learning environment or a physical classroom, a single facilitator or teacher will find it impossible to divvy up her attention and time among all learners to offer the in-depth formative assessment and feedback they need. Enter the AI virtual assistant, which can be trained based on relevant data sets to assess learners in real-time and guide them to improvement with coaching and questioning. Different users will perform differently and require different layers of feedback to improve. This also keeps the user constantly motivated and engaged - there is no time lag between their effort and assessment, which charts them to stay on course.

Guardrails and Use Cases

Evidently, you would need a colossal bank of relevant, accurate and meaningful data in order to train chatbots and AI tutors which are accurate, safe and useful. You may also need a significant investment of time, money and effort to build AI chatbots, skills coaches and intelligent support systems which are reliable and accurate. However, the barriers to entry may be lower than you think, particularly due to the number of AI tools in the marketplace and the advancements that are happening. Most importantly, you will be able to build on prior use cases and knowledge, lowering the costs and time to market for your AI solutions.