Thanks! I was reading a LinkedIn post just this week where someone was talking about a Turnitin portal, and they were going to require students to essentially log into it to do their work while everything they did in the writing process was recorded. They would then use an AI driven AI detector via Turnitin to determine if students wrote the work themselves.
I've watched a lot of evaluations on these AI detectors and they are wildly different since they are run by proprietary organizations. I also haven't seen what the actual standard is from any type of educational authority on what is an acceptable AI or human score. What I do see is random educators applying their own standards, using free AI detectors or sneaky things like invisible texts embedded into assignments or a syllabus, and then holding students accountable for AI driven work when no one has actually been trained on how to create responsible AI content.
We've got teachers out here using free tools, students using free tools, and for profit organizations trying to swoop in to solve a problem that hasn't been properly addressed in the first place. non-English speakers, minorities, and others who may simply try to use AI as a writing coach are getting falsely accused of writing the whole thing in AI. As you can see from my article, parents and others are beginning to get fed up and are starting to file lawsuits over it.
The AI developers don't get off scott free here either though because they're truly isn't a free tool available to the public to use for training, and I don't know how fast people can really learn with things like 10 prompt a day limits.
Thank you for this- I now understand better why that article I genuinely wrote - was classed as written by AI by a publication (when in fact it was not). I had used Grammarly - to help with spelling and grammar checks. I think this was it. It would be interesting to also hear about authors who write AI-generated books, what are the risks, can an AI company sue them for this?
I personally I am not against writing with AI as long as the person writes the article and polishes / edits it with the help of AI. We should not call ourselves writers if all we write is generated by AI.
There is nothing wrong with using AI as an editor or proofreader, and I feel like it's no different than the grammar corrector in Microsoft Office or Grammarly. The only time you're really going to get in trouble with copyrights, based on current rulings by the copyright offices, is if you are literally taking a copyrighted piece of work and using AI to create a virtual duplicate of it. The copyright office recently ruled that AI created works that were created with a significant degree of human input can be copyrighted, but content created by AI tools that severely limit or prohibit human input in the creation process can't be copyrighted.
I'm sure we're going to have to see more use cases and litigation on the matter before we get a real concrete standard that everyone can understand.
very interesting, and I say this because AI is getting better and better with even mimicking human emotion in writing. While in the past it used to be more rigid and formal, now it is learning from our interactions with it and adopting its voice. For example, I usually like to add a fun tone of voice in my writing, and when I ask Chat GPT to polish my article it usually tends to amplify my tone of voice. So, naturally, I am curious to see where this is going because it will get harder to distinguish between what's written by AI and what's written by humans.
Great article! It seems to me that we (educators) need to rethink how we build learning activities and assessments so that they either make it difficult or unproductive to use AI, or that leverage AI. Currently, trying to detect AI writing is exhausting.
Great article! It seems to me that we (educators) need to rethink how we build learning activities and assessments so that they either make it difficult or unproductive to use AI, or that leverage AI. Currently, trying to detect AI writing is exhausting.
Yes I think that is a good suggestion. I would encourage lessons on assignments that leverage AI because we need our future workforce to be able to use it, and it's an opportunity to teach responsible AI use at the same time.
The workforce aspect is huge. I'm involved with a couple of industry groups and they absolutely want new hires to understand at least the basics of generative AI. In fact, some colleagues and I are slowly trying to nail down exactly what sorts of generative AI skills employers seek. My initial sense is that they don't really know.
Such a great post! I appreciate that you addressed the ethical implications too. Thank you for sharing Robert!
Thanks! I was reading a LinkedIn post just this week where someone was talking about a Turnitin portal, and they were going to require students to essentially log into it to do their work while everything they did in the writing process was recorded. They would then use an AI driven AI detector via Turnitin to determine if students wrote the work themselves.
I've watched a lot of evaluations on these AI detectors and they are wildly different since they are run by proprietary organizations. I also haven't seen what the actual standard is from any type of educational authority on what is an acceptable AI or human score. What I do see is random educators applying their own standards, using free AI detectors or sneaky things like invisible texts embedded into assignments or a syllabus, and then holding students accountable for AI driven work when no one has actually been trained on how to create responsible AI content.
We've got teachers out here using free tools, students using free tools, and for profit organizations trying to swoop in to solve a problem that hasn't been properly addressed in the first place. non-English speakers, minorities, and others who may simply try to use AI as a writing coach are getting falsely accused of writing the whole thing in AI. As you can see from my article, parents and others are beginning to get fed up and are starting to file lawsuits over it.
The AI developers don't get off scott free here either though because they're truly isn't a free tool available to the public to use for training, and I don't know how fast people can really learn with things like 10 prompt a day limits.
Thank you for this- I now understand better why that article I genuinely wrote - was classed as written by AI by a publication (when in fact it was not). I had used Grammarly - to help with spelling and grammar checks. I think this was it. It would be interesting to also hear about authors who write AI-generated books, what are the risks, can an AI company sue them for this?
I personally I am not against writing with AI as long as the person writes the article and polishes / edits it with the help of AI. We should not call ourselves writers if all we write is generated by AI.
Lexy,
There is nothing wrong with using AI as an editor or proofreader, and I feel like it's no different than the grammar corrector in Microsoft Office or Grammarly. The only time you're really going to get in trouble with copyrights, based on current rulings by the copyright offices, is if you are literally taking a copyrighted piece of work and using AI to create a virtual duplicate of it. The copyright office recently ruled that AI created works that were created with a significant degree of human input can be copyrighted, but content created by AI tools that severely limit or prohibit human input in the creation process can't be copyrighted.
I'm sure we're going to have to see more use cases and litigation on the matter before we get a real concrete standard that everyone can understand.
very interesting, and I say this because AI is getting better and better with even mimicking human emotion in writing. While in the past it used to be more rigid and formal, now it is learning from our interactions with it and adopting its voice. For example, I usually like to add a fun tone of voice in my writing, and when I ask Chat GPT to polish my article it usually tends to amplify my tone of voice. So, naturally, I am curious to see where this is going because it will get harder to distinguish between what's written by AI and what's written by humans.
Great article! It seems to me that we (educators) need to rethink how we build learning activities and assessments so that they either make it difficult or unproductive to use AI, or that leverage AI. Currently, trying to detect AI writing is exhausting.
Great article! It seems to me that we (educators) need to rethink how we build learning activities and assessments so that they either make it difficult or unproductive to use AI, or that leverage AI. Currently, trying to detect AI writing is exhausting.
Yes I think that is a good suggestion. I would encourage lessons on assignments that leverage AI because we need our future workforce to be able to use it, and it's an opportunity to teach responsible AI use at the same time.
The workforce aspect is huge. I'm involved with a couple of industry groups and they absolutely want new hires to understand at least the basics of generative AI. In fact, some colleagues and I are slowly trying to nail down exactly what sorts of generative AI skills employers seek. My initial sense is that they don't really know.