AI detection software is horribly inaccurate. It thinks the US constitution and passages from the Bible are AI generated. It makes a guess that spits out a percentage and an unreliable one at that. And yet it's being used to accuse kids who did nothing wrong.
In contrast, plagiarism detection software shows an actual comparison between the writing being looked at and any other existing content that is similar. It allows for the teacher to make there own assessment. AI detection software should be banned -- the risk of unfairly disqualifying an entire swatrh of applicants is not worth the odds of possibly catching someone who actually use AI to do all their work. Sadly, teachers and institutions who use AI detection software are likely to be too complacent with being able to outsource their responsibility for evaluating students, and the companies who sell them the software have their business model at stake and will be very difficult to fight against. Turn-it-in has already had to put our revisions to its earlier claims of a 4 percent error rate which provide to be completely off. Chat-GPT actually terminated their detection software because of its unreliability. And agree with others that the irony is that it will be the privileged families and their kids who can pre-scan their work through detection software algorithms and then revise it to avoid getting caught. It will be the disadvantaged families who will suffer. |
the degrees that focus on literature don't prepare you for a barista so its useless, career options are a professor of literature or entry level food service field |
wow, I wonder how long until a student accused by turnitin sues |
People are pretty bad at telling if AI wrote something. Remember the guy who said that if “delve” was used, it’s a sure sign that it was AI??
He also listed a few other words that I’ve definitely used. Maybe I’m a computer. |
Writing is thinking. If a student has not learned to create fluid prose on his own, he will be useless at creating good content with AI, not to mention not capable of discerning truth or BS fed to him by a tech company. This is why IQs are going down… |
We can tell… |
Do you have kids in college? Most professors have had to adopt a lot of high school methods to try to force kids to read, show up for class and chunk projects. |
The child is a junior… the teacher should do this, not mom. |
All the teachers, including so-called college professors, coming on here saying the solution is to demand the kids rewrite their papers or outlines from memory are completely out of touch. In a million years, I could never rewrite an outline or paper that I had spent days and days on from memory. Both my college kids have ADHD, they have to spend twice as long outputting excellent papers as regular students do. And when they're fininished, they can barely tell you what they wrote--certainly not from memory. It takes hous for them to fill in very detailed outlines, then turn them into well-written papers, which they write and re-write over and over. And yes, both of them will write a sentence, then write into Chat GPT or run it through Grammarly, with a prompt that says: fix this for me. |
Lol |
I work in a strategy field and use AI daily for work. I’ve found that if I ask ChatGPT to “rewrite for clarity and/or simplicity,” I get usable content. I will also add an example ofmy own writing style and will ask it to copy my style.
Here’s ChatGPT’s written for “clarity” example: We should support using AI in schools to help with tasks like grammar and organizing essays. Instead of just focusing on those details, students should concentrate on the main ideas and themes in their writing. They should present their work and talk about it to show they understand the material. Instead of punishing students for using AI, we should encourage it. Soon, being good at using AI will be important for grading, based on how well students use it to improve their work. |
Why Aren't school teaching students how to use ai for their essays and work? |
I’m a high school teacher. I require Google docs for all assignments. I don’t need to use AI detectors because the version history and draftback give me all the proof I need. When two well-constructed paragraphs are recorded as one edit, I have my proof it was a cut/paste job. Students will often say they wrote those paragraphs in another document, which is against policy for this very reason. 100% of the time, when I ask for access to this new doc, they can’t provide it. We use chatGPT in class to explore its functions and to generate ideas, but we also spent two full days exploring how it should/should not be used in an academic environment. We looked at AP and IB and collegiate policies on AI. All of this is clearly outlined in my syllabus, as well. |
Can you take the formulaic piece and make it your own so it’s not so obvious? |