Better Late than Never…

Waiting by Edgar DegasPublic Domain

The Journey of a Thousand Miles Begins With a 3 Week Late Blog Post

September is busy time in the life of the teacher.  Classes need prepping, coffee needs consuming, and major projects on Artificial Intelligence (AI) need researching.  Two of these things began in earnest over a month ago, and the other didn’t get a serious start until this long weekend.

Welcome to the blog where I intend to regale you with tales of my progress toward the completion of my EC&I 832 major project: Exploring Generative AI and Ethical Digital Citizenship.  So what have I been up to so far?  Mostly stumbling through academic readings while I try to wrap my head around the topic.  However some insights have been gained and interesting articles found.

AI and the Classroom – Starting With What I Know

When I selected this project I thought it would be interesting for two reasons:

  1. I have never used AI based tools in my classroom and know literally nothing about them (except that English teachers in my school have developed a nervous tick since they’ve become widely available).
  2. Everyone that I’ve talked to about them regards with them with suspicion – this tells me that they are more than a “flash in the pan” as fundamental shifts in society are often met with fear.

As a teacher my first inclination was to go with what I know the classroom.  How are people using AI productively?  What can it do and not do?

The first article I read by Tyson and Sauers gives some insight into how administrators in schools view AI technology and details how it might be used productively to reduce teacher workloads.  Two tidbits that seemed particularly interesting were using the technology to help develop IIPs (individual instruction plans) and to reduce “busy work” by creating letters home, permission slips, and other correspondence.  Further to this they emphasized how important clear communication with the community was to the successful deployment of AI based tools.

Another article I read discussed how to integrate Chat GPT into classrooms by using it as “colleague” to bounce ideas off of.  The author, Ashley Abramson, suggested that AI tools like chat bots would not be appropriate for teaching fundamental skills (like grammar), but instead frees up mental resources for higher level discussion (much like calculators in a calculus class).  Interacting with AI allows students to think critically like they would in any other conversation – not everyone is an expert in all things and need to be challenged on the merit of their arguments.

A interesting thread gets pulled…

After looking at a few more articles centered on the educational applications on AI I stumbled across an article delving into the bias reflected in (and amplified by) machine learning and intelligent systems.

In it the authors argue that since AI systems train themselves on large data sets that are ridden with potential biases the conclusions that they draw exhibit the same (and sometimes increased) level of bias.  Several examples were given including AI systems denying people of colour bank loans, restricting medical treatments based on race and ethnicity, and court utilized algorithms disproportionately suggesting harsher sentencing and risk to reoffend based on race.  The article then when on to detail the challenges of eliminating bias in algorithms focusing on 3 approaches (pre, during, and post-processing).  While some of the language was over my head – I believe this could be an interesting angle with which to approach my project.  I intend to look deeper into this over the course of the next week.

Leave a Reply

Your email address will not be published. Required fields are marked *