I never use AI for research or for writing in any way and do not anticipate ever doing so, for all the reasons you discuss and then some. I cherish my own cognitive development and my carefully honed judgment and evaluative abilities as well as my choice to constantly work on honing my thinking and writing skills. For goodness sake, I refuse to use spellcheck or grammarcheck, both of which have flaws. I do my own editing and proofreading based on my knowledge (which I also work hard to refine), so any remaining errors of style, punctuation, and grammar are entirely my own. Using AI would only rob me of all that. Perhaps I should state on my newsletter that *that* is my policy: no AI at all.
I’m among those who use AI as a tool…best described as a brainstorming partner (because I am a solopreneur and other wise have no one to bounce ideas around with) to ensure I haven’t overlooked anything and to suggest other angles or way to improve something I’ve written myself. I have been frustrated with ChatGPT too and am preferring Perplexity.ai because it provides citations for everything. I appreciate your insight and you’ve made a great case for having an AI policy.
Great article Karen. I have an AI usage policy on my Parents who Think substack. I wrote it without AI and it was triggered by the upcoming work I’m doing on how AI might transform unequal distribution of labour in homes. In covering AI I realised I needed to be transparent about whether I use it for writing. Plus how I use it within our home. Love that you’re advocating a policy on this. It’s becoming more important, right?!
I’ve considered writing an AI policy and now I think I may actually do it. Your thoughts about the way AI can waste time instead of saving it make a lot of sense also. :)
HI Karen. I have absolutely no interest in using AI at all. I actually enjoy researching for my newsletter and books, I love discovering and appreciating the work of others that I can learn and cite from, and the research rabbit holes I often fall down. I feel satisfaction from creating my own documents, I would never think of getting AI to generate anything for me, that would not be rewarding to me. But... I do take all my own photos for the newsletter on my smartphone, and I use the camera's inbuilt editing facility, so I guess that is where I am using AI. It's almost impossible to escape it completely!
This is excellent, Karen! Because of the type of content I publish in my newsletter here, I doubt I’ll create an AI policy page, but I can see why you would want to. I appreciate your honesty and integrity. See, THIS is truly what’s missing in the evolving landscape of writing: people don’t care if AI is accurate or not. It sounds good. It’s fast. Everyone is using it, so might as well jump on board.
Nope! Not me, either! I use chatGPT for help with formatting tricky citations, and last week I used it to help me generate and organize commonplace book sections and templates for my internal note-taking system.
My husband started using Google NotebookLM recently and I’m considering it, too. Unlike ChatGPT, YOU have to provide it the sources. I use Logos Bible Software and it works in similar fashion. It’s mostly useful for providing summaries, searching, or coming up with questions. The latter helps me think.
But I do NOT use AI for research for all the reasons you mentioned. I also don’t use AI to generate any part of what I write other than occasional re-writes of sentences (for clarity) by the free version of Grammarly.
There you go. I just wrote my AI policy (more or less) didn’t I? 😄😎
I don't know. I understand why so many people—writers mostly—fear AI. And it's very popular to hate on it these days. But I view AI in all its forms as a tool.
Would we even dream of explaining to people that we use spell check or grammar check within Microsoft Word? Why then should someone feel they have to explain that they run their work through a tool such as Grammarly?
If we search for a fact in ChatGPT instead of Google is that somehow more relevant? We have to verify it either way. Junk facts on the internet are not exactly new.
Again, I get it. But AI is now, or soon will be, built into practically every device or software that we use. How far do we take it? It's not just ChatGPT.
Ultimately, the quality of your work is what truly matters. If you're relying solely on AI to create it will likely show (at least for now). But if you're using it as a tool to enhance your existing skills and the results are effective, why should it matter how you got there?
I used AI to "enhance" one of the above paragraphs. Which one? Does it detract from my point in any way?
Jason, I take your points, but understand the nature of the quality of my work somewhat differently. It's not merely informational output as though an author is like a calculator that spits out answers to an equation. There are also issues of judgment, style, trust, and authenticity that get inflected in and communicated by the content that only you can give your expression to as a reflection of your understanding of something. By the way, I never use the spell check or grammar check in Word, for the reasons I just gave in my comment on Karen's post. Thank you for your "food for thought," which has provoked me to think harder about this issue!
Maybe it would help if you knew more about AI before opining a best practice? Going by what you have written you are basing your knowledge on what you can achieve as a user.
There are people using AI who actually know what they are doing and keeping up with the industry. Some of them have Substack publications. Just sayin'.
It feels weird that we have to talk about this lol. My AI page would be pretty short. I use Grammarly and that's it. I don't think anyone cares about that and half the time it doesn't recognize my fun personality so I ignored its suggestions lol.
Karen, I love that you’re advocating for writers to be transparent about writing down & sharing their AI usage policy. As you point out, everyone has one - it just may not be written down yet 😊
It's helpful to know how you're providing transparency to your readers, Karen. While I may not produce an AI policy for my Substack work, I'm considering it for my website, where I currently house most of my writing. Interestingly, I self-published my first book last fall and have had several industry professionals ask me if I wrote it. (YES!) Ghostwriting has been standard practice in the publishing world....yet there's no ethical code on that (did you know it's estimated 50%+ of business books are ghostwritten??? I didn't!)
I know a few people who have used AI to write content like biographies, personal profiles, etc. AI uses certain phrasing that makes this content sound the same to my "ears." I think AI can be useful for proofing and research. Possibly for brainstorming ideas. But I have no interest in using AI to write my content. My writing voice is such an important part of the way I develop trust-based relationships with my readers. I want my work to sounds as authentic as possible. Why would I want to dilute it?
AI is a tool to assist with creating not a replacement.
I never use AI for research or for writing in any way and do not anticipate ever doing so, for all the reasons you discuss and then some. I cherish my own cognitive development and my carefully honed judgment and evaluative abilities as well as my choice to constantly work on honing my thinking and writing skills. For goodness sake, I refuse to use spellcheck or grammarcheck, both of which have flaws. I do my own editing and proofreading based on my knowledge (which I also work hard to refine), so any remaining errors of style, punctuation, and grammar are entirely my own. Using AI would only rob me of all that. Perhaps I should state on my newsletter that *that* is my policy: no AI at all.
I’m among those who use AI as a tool…best described as a brainstorming partner (because I am a solopreneur and other wise have no one to bounce ideas around with) to ensure I haven’t overlooked anything and to suggest other angles or way to improve something I’ve written myself. I have been frustrated with ChatGPT too and am preferring Perplexity.ai because it provides citations for everything. I appreciate your insight and you’ve made a great case for having an AI policy.
Thanks, Elizabeth, I’m also a fan of Perplexity.
Great article Karen. I have an AI usage policy on my Parents who Think substack. I wrote it without AI and it was triggered by the upcoming work I’m doing on how AI might transform unequal distribution of labour in homes. In covering AI I realised I needed to be transparent about whether I use it for writing. Plus how I use it within our home. Love that you’re advocating a policy on this. It’s becoming more important, right?!
I’ve considered writing an AI policy and now I think I may actually do it. Your thoughts about the way AI can waste time instead of saving it make a lot of sense also. :)
HI Karen. I have absolutely no interest in using AI at all. I actually enjoy researching for my newsletter and books, I love discovering and appreciating the work of others that I can learn and cite from, and the research rabbit holes I often fall down. I feel satisfaction from creating my own documents, I would never think of getting AI to generate anything for me, that would not be rewarding to me. But... I do take all my own photos for the newsletter on my smartphone, and I use the camera's inbuilt editing facility, so I guess that is where I am using AI. It's almost impossible to escape it completely!
This is excellent, Karen! Because of the type of content I publish in my newsletter here, I doubt I’ll create an AI policy page, but I can see why you would want to. I appreciate your honesty and integrity. See, THIS is truly what’s missing in the evolving landscape of writing: people don’t care if AI is accurate or not. It sounds good. It’s fast. Everyone is using it, so might as well jump on board.
Nope! Not me, either! I use chatGPT for help with formatting tricky citations, and last week I used it to help me generate and organize commonplace book sections and templates for my internal note-taking system.
My husband started using Google NotebookLM recently and I’m considering it, too. Unlike ChatGPT, YOU have to provide it the sources. I use Logos Bible Software and it works in similar fashion. It’s mostly useful for providing summaries, searching, or coming up with questions. The latter helps me think.
But I do NOT use AI for research for all the reasons you mentioned. I also don’t use AI to generate any part of what I write other than occasional re-writes of sentences (for clarity) by the free version of Grammarly.
There you go. I just wrote my AI policy (more or less) didn’t I? 😄😎
I reckon you did, Sharon. Job done!
Here’s my AI policy. I don’t use it, and I will not support any writers who use it.
I don't know. I understand why so many people—writers mostly—fear AI. And it's very popular to hate on it these days. But I view AI in all its forms as a tool.
Would we even dream of explaining to people that we use spell check or grammar check within Microsoft Word? Why then should someone feel they have to explain that they run their work through a tool such as Grammarly?
If we search for a fact in ChatGPT instead of Google is that somehow more relevant? We have to verify it either way. Junk facts on the internet are not exactly new.
Again, I get it. But AI is now, or soon will be, built into practically every device or software that we use. How far do we take it? It's not just ChatGPT.
Ultimately, the quality of your work is what truly matters. If you're relying solely on AI to create it will likely show (at least for now). But if you're using it as a tool to enhance your existing skills and the results are effective, why should it matter how you got there?
I used AI to "enhance" one of the above paragraphs. Which one? Does it detract from my point in any way?
Just food for thought.
Jason, I take your points, but understand the nature of the quality of my work somewhat differently. It's not merely informational output as though an author is like a calculator that spits out answers to an equation. There are also issues of judgment, style, trust, and authenticity that get inflected in and communicated by the content that only you can give your expression to as a reflection of your understanding of something. By the way, I never use the spell check or grammar check in Word, for the reasons I just gave in my comment on Karen's post. Thank you for your "food for thought," which has provoked me to think harder about this issue!
Maybe it would help if you knew more about AI before opining a best practice? Going by what you have written you are basing your knowledge on what you can achieve as a user.
There are people using AI who actually know what they are doing and keeping up with the industry. Some of them have Substack publications. Just sayin'.
It feels weird that we have to talk about this lol. My AI page would be pretty short. I use Grammarly and that's it. I don't think anyone cares about that and half the time it doesn't recognize my fun personality so I ignored its suggestions lol.
haha yes Kristi, I derive weirdly intense pleasure from ignoring Grammarly suggestions!
Karen, I love that you’re advocating for writers to be transparent about writing down & sharing their AI usage policy. As you point out, everyone has one - it just may not be written down yet 😊
It's helpful to know how you're providing transparency to your readers, Karen. While I may not produce an AI policy for my Substack work, I'm considering it for my website, where I currently house most of my writing. Interestingly, I self-published my first book last fall and have had several industry professionals ask me if I wrote it. (YES!) Ghostwriting has been standard practice in the publishing world....yet there's no ethical code on that (did you know it's estimated 50%+ of business books are ghostwritten??? I didn't!)
I know a few people who have used AI to write content like biographies, personal profiles, etc. AI uses certain phrasing that makes this content sound the same to my "ears." I think AI can be useful for proofing and research. Possibly for brainstorming ideas. But I have no interest in using AI to write my content. My writing voice is such an important part of the way I develop trust-based relationships with my readers. I want my work to sounds as authentic as possible. Why would I want to dilute it?
AI is a tool to assist with creating not a replacement.