By Elijah Horman
A new threat looms over higher education institutes, threatening students’ academic and creative integrity. New software also threatens all walks of life from software engineering to public policy.
Launched on Nov. 30, ChatGPT was created by OpenAI, and is capable of writing essays, code, stories, lesson plans and many other forms of writing. Within its first week of launch ChatGPT already crossed over 1 million users.
Education institutes have started becoming wary of these new programs, causing instructors to rethink traditional writing assignments.
Professor Caleb Husmann is the assistant Political Science Professor here at WPU who has done extensive research into ChatGPT and is beginning to adapt his class to the new program.
“I’m also doing far more writing in class, adjusting to make some things like a debate or a presentation or something of that nature as opposed to what used to be like a short paper,” said Husmann.
In person, and oral assignments are a definitive way to negate the use of ChatGPT allowing for professors to more easily see how students are writing.
There are, however, more limitations to the software in its current state. The data the ChatGPT pulls from is from 2021 and before, so it will struggle to piece together data from more recent events.
“The more contemporary it is, like the more recent, the assignment the better…” said Husmann “so if I asked you to write an assignment based off something that’s in the news today, it’s not going to do as good of a job.”
In fact, according to Husmann, the two types of assignments that are in the most danger are short creative stories, and summary assignments.
Short stories are under threat because ChatGPT can write in the style of any author published online. The only saving grace to this is that currently ChatGPT has a 650 word cap on any prompts, though users are already finding ways around this limitation.
Summary assignments are under threat for much the same reason. So long as the book or article was published on or before 2021, the AI will be able to summarize it.
ChatGPT and AI writers as a whole are advancing at a formidable rate. Currently the most recent version is version 4, but openAI is working quickly to release new updates. This may include a greater selection of data including more recent events and an increased number of parameters which will lead to more accurate responses.
In addition to this there is a version of the ChatGPT system released as open source. While currently it is fairly barebones and difficult to run, opening this system up to the general public will lead to significantly faster advancements.
Schools across all levels have already begun to adapt their curriculums to fit with the new program. Some middle and high school level institutes are testing an outright ban on the program, however, most higher level institutes have not yet taken this step as it would most likely prove ineffective.
Higher education institutes have instead begun to phase out take-home essay assignments in exchange for more oral/presentation assignments, or even simply requiring that the first draft of an essay be written in class where the computer can be properly monitored.
Luckily for educators OpenAI has begun to design a watermark system put into all works created by ChatGPT. These kinds of watermarks are often seen in art as a semi transparent logo or design signifying where the work came from. For ChatGPT this would be a cryptographic code embedded into the words, letters, and punctuation that anti plagiarism software could more easily pick up on.
ChatGPT, and other AI writers/drawers pose a new challenge to our education systems, and without our schools adapting as well as OpenAI inputting anti plagiarism measures these writers pose a massive threat to both academic integrity, and student creativity.