AI in Education (part 3)
In the previous two blog posts (part1 and part2), I mused about the awesome potential of AI in education. In particular, Machine Learning (ML) is a mapping technique that can make use of the huge availability of digital content for training. ML models can be trained for various subjects, grade levels, and (eventually) the learning preferences of individual students. While the initial ML model will not have student-specific data, over time, the model can be retrained to include student data.
The potential of AI and ML will likely run into resistance as it faces the same skepticism that EdTech tools have generally faced. This includes the following challenges: (1) tougher to manage students using devices, (2) potential time taken away from the lesson plan for logins and troubleshooting, (3) view that technology is not a critical component of master teaching, and (4) difficulty recreating the lesson plan to use technology in a meaningful way. Teachers will need to gain comfort with AI/ML much in the same way that they built up confidence using EdTech tools. This comfort will likely come in phases (see below). And, let’s not forget the role of superintendents, principals, and curriculum designers. These folks will likely be gatekeepers once AI/ML takes strides toward significantly disrupting learning.