There is a lot of talk about LLMs and their impact on education. Here is a recent article about the potential impact for standardized assessments: https://lnkd.in/grVE37fn In our work, we subdivide the use of AI through a framework that considers education as an embedded system. Within this framework, we distinguish between strategy and tactics (see chess research for AI). Quickly here is a definition: -Strategy: Strategy in the classroom involves long-term planning and the overall approach to teaching, focusing on broader educational goals, curriculum design, and creating a conducive learning environment. It's about setting up the right learning conditions and overall learning trajectory for students over time. -Tactics: Tactics are short-term actions, techniques, or teaching decisions made to address specific learning moments or challenges. They are about immediate responses to student needs, leveraging "teachable moments," and using concrete instructional techniques to achieve particular learning objectives. Humans tend to be better at strategy through intuition and LLMs (AI) tend to be better at tactics. When they are combined effectively, a simple computer and a strong chess player can beat even the most powerful AIs. What would this do in education? Where would you categorize standardized assessments between tactic and strategy?
Jason Beutler Kendra Beutler would love your thoughts here too!
The analogy of chess is apt, as history shows us that technology often excels at tactical execution while humans retain an edge in strategic foresight. Think about the Industrial Revolution machines revolutionized production , but human managers still set the overall direction . Where standardized assessments fall within this framework is a fascinating question. Do they primarily measure immediate knowledge recall or gauge a student's ability to apply knowledge in complex, evolving situations ?
I find personalized AI tutors incredibly fascinating. I’ve been working on one for language learning too. It still needs some fine-tuning, but it’s getting closer every day. Dr. Kevin Berkopes
It’s almost like you are reading the book Range… the idea of humans being particularly skilled at strategizing in the wicked world while computers take care of the tactical necessities is really compelling
CEO and Cofounder | MathTrack Institute: Work-based training and development for mathematics
5moTyler Foxworthy 🤙Gleb Alexeev Further, could standarized assessments become longitudinal data streams where a student's proficiency is measured over time with adjustments comparative to the data rather than the reductive approach today of "snap shots"?