We can now identify AI writing

Many of us have been reading about and commenting on the use of AI systems to produce written texts. Like some others, I’m worried about the impact this could have in education where students may be tempted to use AI to write their assessments. Because AI will produce their own words there is little possibility of plagiarism and as such, it would be difficult to detect when AI is being used.

I read an article today about a student in the US who has developed software to detect when AI is being used to write rather than human-composed text. I found that article hopeful and promising.

I’m sharing a copy for you to read too.

https://www.smithsonianmag.com/smart-news/student-creates-app-to-detect-essays-written-by-ai-180981463/

I’d love to read your thoughts in the comments section.

Subscribe to my blog for notification of my future posts.

Advertisement

33 comments

  1. 🤔 I have dealt with the topic of ChatGPT a few times on my blog.

    I will admit that it is a cause for concern. After all, it takes away the ability of students to think critically and come up with their own answers.

    GPTZero does a wonderful job of detecting AI-generated content.

    At the moment, there are some young geniuses who are working on a way to fool GTPzero and other software that can detect articles that were generated by artificial intelligence (I have a feeling that they are going to be successful at it).

    Think of it as a game of cat and mouse; today students’ assignments can be determined if it was created by them or AI and tomorrow, the detection is bypassed (despite using an AI writer) and the assignment is categorized as being, “100% real.”

    Liked by 4 people

    • Oh I agree totally Renard. Cat and mouse is the exact way to refer to it,and we’ll always be playing catch up. Fortunately for some of the work I’m dealing with students need to apply it to their own employing organisations, so that might make it more challenging for AI. However, I agree that AI will adapt and move forward. We will always be playing catch up.

      Also, I guess some students will always look for ways to cheat. This is simply the most recent version of that.

      Liked by 2 people

    • Youre welcome Wynne. I’ve come across a couple of articles in the teaching press that I’ll need to check out … but I’ve a pile of marking to get through first.

      Liked by 2 people

  2. An interesting article you shared, Brenda. As time moves on, the use of AI will develop and become more efficient. In the future I would like to think AI will be used to support writers rather than replace them. I have tried a couple of AI apps to see how they write poetry. They had a kind of Spike Milligan feel about them. Writers will always write because of the desire they have inside them. I do agree with you about the concerns for education. I remember a teacher having similar concerns in the early 1970s when someone bought me a Texas Instrument Mathematical Calculator. (I’m still bad at mathematics.)

    Liked by 3 people

    • I can relate to the maths thing … numbers 😵‍💫
      I think we can use tools but we need to know the basics ourselves … you need to beable to tell if the calculators answer seems real/correct.

      I’ve been marking today and I can see that because a student doesn’t know what references are supposed to look like, they can’t tell the layout is wrong … guess the old adage still holds true – the tools are only as good as the operator

      Liked by 1 person

  3. Happy, really happy to find “some people” who think the thoughts in my heart.

    I am in no way against the use of Generative AI [for text and art]. My concern is that, like every other good thing man has created, we will abuse AI.

    AI is supposed to help our creativity help us generate ideas faster while saving us a lot of time. But with the way things are going, we might end up depending too much on it to the extent that if we do not have access to our “AI Toys,” we may not be able to come up with even a paragraph.

    Though the news of folks coming up with ways of detecting AI-generated texts [and images] is somewhat promising, we still have to tackle the problem of folks trying to bypass the “AI Police.”

    It’s, as said, an interesting game of “cat and mouse.”

    Thank you for the share, Brenda.

    Liked by 2 people

    • Thank you for reading and commenting. It is a complicated issue. I’ve just been reading and marking some student work where they are exploring what technology will be like in the future. They came up with some interesting ideas.

      Liked by 1 person

  4. Remarkable blog Brenda. This All writing is taking the world by storm hey, well I really don’t like the way Al apps such as Writersonic are doing all the writing now and for the students they will rely on Al to write their assignments and then they will submit claiming its their own work. This Al will make people lazy to write and think in the process as well💯👏

    Liked by 2 people

    • It is definitely a risk, and it will encourage those who are already lazy or struggling to take the easy option. I was worried about essay writing services,but I think there’s a new worry 😟
      Things are always developing and adapting. It’s the nature of life I guess

      Liked by 2 people

  5. Since the accuracy rate for the latest ChatAI is in the 80%s, teachers can adapt by changing assignments to include more critical thinking and specific facts. It won’t be able to cite sources it used either.

    Liked by 2 people

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s