GAiIT Level 2: When Students Cross the Line
It now IS happening!
Much of the Ai abuse stems from the same sorts of motivations. Only now, detecting it became far more nuanced and tricky to find. Use the example we provided for the G2 example to identify similar issues with your own G2 level assessments.
A lack of clarity for the expectations of the use of Ai
A lack of time management leads to use of Ai doing the work
A lack of subject investment leads to Ai doing the work
A lack of subject knowledge leads to Ai doing the work
A lack of belief that the instructor will read their work
A lack of technology training that would prohibit the student from completing the work
What to look for at GAiIT Level 2
The Ai has officially be invited in as a co-creator and collaborator in the assessment. Essentially, the Ai will be the workhorse and the student would absorb a more managerial role evaluating and revising the work produce by the Ai towards any given project. In the example provided for GAiIT Level 2, we watched the student undertake the enormous task of building an entire e-magazine on their own. Well, not entirely on their own. Of the 10 articles Ai would write 7 of them and potentially re-write three of the student's more academic reading pieces. Although the Ai is taking on a substantial amount of the work, the student has to process the output ensure that the e-zine (or other project) emerges with both its written and visual content tightly arranged around the content of the article.
When looking for Ai in writing assessments:
The e-zine assignment is supposed to sound conversational and human-like which will require a fair amount of editing from the student. Magazines very rarely communicate like white-papers, and Ai tends to communicate like that in written form we might be looking out for some of the same things:
- Mechanical and impersonal opening statements
- Short brief paragraphs without using targeted cited sources
- Listing of content in paragraph form
- Speaking in third person when the object of the lesson is to have a first person reflection
- Overly mechanical and formal writing free of linguistic euphemisms used by native speaker'
- The use of the phrase, "In conclusion,"
When looking for Ai in project-based assessments:
- NOT CITING THE WORK ATTRIBUTED TO THE Ai = #1 offence!! It was a collaborator; like any collaborator, their work should be noticed and acknowledged.
- Canned titles - Magazine and non-academic publications use double-entendres and puns in titles. A noted lack of these might indicate Ai made the title.
- Images that seem out of touch from the article - Ai image generators use a prompt to signal them what to do, without specific prompts the Ai will fill in blanks with its own ideas. Often, these ideas are a little off. Most of the images for a targeted project would require several revisions before the image looks right for the written content.
