A novel AI-powered application called Writable, designed to assist teachers in grading student writing assignments by leveraging ChatGPT’s capabilities, is gaining widespread adoption among educators teaching grades 3 through 12.
The Significance of AI in Grading
While educators have been discreetly employing ChatGPT to evaluate papers since its inception, the use of such AI tools is now being sanctioned and actively encouraged by educational institutions.
Writable’s Acquisition and Widespread Use
Writable, touted as a time-saving solution for teachers, was acquired last summer by the educational publishing giant Houghton Mifflin Harcourt, whose materials are utilized in 90% of K-12 schools across the nation. [1] Numerous teachers are enthusiastically embracing this program, as stated by Jack Lynch, the CEO of Houghton Mifflin Harcourt.
How Writable Operates
- Teachers assign a writing task, and students electronically submit their essays.
- The teacher uploads the essays to Writable, which then processes them through ChatGPT.
- ChatGPT provides feedback and comments, which the teacher reviews and refines before sending back to the students. [3]
- Writable “tokenizes” student information to ensure privacy during the AI evaluation process.
The “Human in the Loop” Concept
According to Lynch, Writable is based on the “human in the loop” AI concept, where the teacher plays a crucial role in reviewing and refining the AI-generated feedback before it reaches the students.
More AI Grading Tools On The Rise
While the implementation of AI in grading practices is not entirely new, a wave of formalized AI grading tools is emerging to institutionalize and codify this approach – a development that carries both potential benefits and drawbacks. Alternatives to Writable in this space include Crowdmark, EssayGrader, Gradescope, and the direct use of ChatGPT, among others. [2] McGraw Hill, another educational publishing giant, has revealed plans to introduce a similar tool that will “use GenAI to provide interactive assessment and feedback to students and insights to educators in writing.”
Concerns and Debates Surrounding AI Grading
Critics argue that AI grading tools could enable teachers to take shortcuts, potentially depriving students of meaningful feedback. While conscientious educators may use ChatGPT’s suggestions as a starting point, others may inadvertently pass along AI-generated feedback verbatim to students.
As schools grapple with formulating AI policies, debates arise regarding the academic integrity of using ChatGPT for grading and the potential impact on student learning experiences. [3]
The Promise of Increased Teacher Productivity
Proponents of AI grading tools, such as Writable, argue that automating grading tasks can free up teachers’ time and flexibility, enabling them to focus on developing creative lessons and fostering stronger connections with their students. According to Jack Lynch, the goal is “to empower teachers, to give them time back to reallocate to higher-impact teaching and learning activities.”
Parental Perspectives on AI-Assisted Grading
While some parents have expressed concerns over AI-generated comments on report cards, a poll conducted by the National Coalition for Public School Options revealed that 45% of parents support the use of AI to evaluate students’ academic performance in K-12 schools.
AI Grading in Higher Education vs. Lower Grades
Historically, AI grading has been more prevalent in large high school and college classes with numerous students, rather than in lower grades where personalized teacher feedback is considered crucial. However, companies like McGraw Hill have been simplifying the grading process significantly through platforms like Connect, which uses non-generative AI to provide auto-scoring and instant feedback based on instructor-set rubrics.
The Necessity of Guardrails and Policies
As ChatGPT and similar AI tools become increasingly integrated into educational settings, the focus should shift toward establishing appropriate guardrails and policies to ensure their responsible and ethical use, rather than dismissing their potential altogether.
