When is it OK to use AI to craft emails, and when do you risk stepping into a whole lot of unexpected hurt? Intrigued by the question, I brushed off my risk hat and dived in.
Long before ChatGPT came out many email messages began with "Please be advised that" included phrases such as "valued customer" and "in this matter" and concluded "and we apologize for any inconvenience caused". These phrases along with other boilerplate that filled some messages are one pre-AI problem.
Spell check replacement errors and other evidence that the message wasn't proofread were also common.
So, while a real, carefully written message is superior to one written by ChatGPT, there are some that don't make the grade.
Thank you! This post provides fantastic perspectives. I recently used ChatGPT to help me craft a message to a mentor on Linked-in....I had a strong emotional reaction to his post and yet I deeply appreciate his ability to ask probing questions. I worked over several days with ChatGPT to consider ways to express my concerns while being respectful of his transparency and experience in expressing probing ideas. So, I expressed these conflicted perspectives to ChatGPT...which provided a wide range of approaches for reaching out to provide thoughtful feedback and to reconnect. I sent a private message rather than responding on the thread..and I was greeted by a positive response. Then my subconscious energy was compelled to share more of my deeply felt concerns and insights...so I responded back from my subconscious instincts. I believe that kindness includes finding ways to share concerns and other perspectives in thoughtful ways....and so I think it's important to consider that genAI systems can be helpful in learning to process and express concerns with those that we respect...so that we're learning to be authentic and helping others expand their world models. In the past, I would send some long emails to my academic peers to see if the ideas resonated with anyone...Now I post comments on substack as a way to explore resonance. When I write substack posts, some are done in collaboration with ChatGPT as I'm trying to weave complex ideas for broad audiences and want to make sure I minimize my personal bias and express the ideas in non-triggering ways. AI tools will continue to improve our ability to communicate complex ideas using powerful examples, just as you have done here. From a modeling and simulation perspective, it's important to learn how to question the results of AI systems and that can result in expanding our world models so that we come to understand we are deeply connected at the level of energy vibrations and that Kindness is the key for learning to improve our relational and transactional communications with others. I also play with using genAI for creating visuals to express the ideas that the universe is fractal energy patterns unfolding and that we can each learn to resonate with these energy fields so we can find ways to be orchestrators of attuned signals that align with sustainable futures for living systems. Keep up the good work!
Really appreciate this post. Like most others reading this, I imagine, I enjoy correspondence and look at it as opportunity for authenticity and rapport—so I write all of mine with no AI.
However, I also try to set aside my own bias as an English teacher and confident writer, as the downsides of a poorly-written, authentic email are heightened when writing isn't a strength for you.
That is where I think this post is most applicable: for those who have to weigh real downsides on either side. Cool approach/lens—much appreciated!
Thanks for this Marcus - one thing I struggled with when writing this was that I know people who genuinely use AI to help with emails as English is a second language to them, or they feel unsure or anxious about their writing ability. Here I do think the tradeoffs between polish and authenticity are interesting -- and not always intuitive.
This is (one of many) reasons I don't use AI to write emails. But, more importantly, I find long emails don't generally get read. If there is something complex that needs to be communicated, I generally go the old fashion way - talk to them in person. Though it's probably what it's used for the least, most of us carry cell phones ...
I write my own emails because relationships matter to me. (And I like writing.)
The only way I've so far figured to message this in-medium (rather than through meta commentary) is to keep email messages so brief that it would be *at least as much* effort to write the prompt, edit, and cut-and-paste.
At that point it becomes credible that I wrote the message myself because it was, in fact, the incentivized option! Of course, this limits the domain of discourse... but I suppose I prefer to use email for shorter communications and, say, a proper briefing note, memo, or position paper, for anything longer.
Long before ChatGPT came out many email messages began with "Please be advised that" included phrases such as "valued customer" and "in this matter" and concluded "and we apologize for any inconvenience caused". These phrases along with other boilerplate that filled some messages are one pre-AI problem.
Spell check replacement errors and other evidence that the message wasn't proofread were also common.
So, while a real, carefully written message is superior to one written by ChatGPT, there are some that don't make the grade.
Thank you! This post provides fantastic perspectives. I recently used ChatGPT to help me craft a message to a mentor on Linked-in....I had a strong emotional reaction to his post and yet I deeply appreciate his ability to ask probing questions. I worked over several days with ChatGPT to consider ways to express my concerns while being respectful of his transparency and experience in expressing probing ideas. So, I expressed these conflicted perspectives to ChatGPT...which provided a wide range of approaches for reaching out to provide thoughtful feedback and to reconnect. I sent a private message rather than responding on the thread..and I was greeted by a positive response. Then my subconscious energy was compelled to share more of my deeply felt concerns and insights...so I responded back from my subconscious instincts. I believe that kindness includes finding ways to share concerns and other perspectives in thoughtful ways....and so I think it's important to consider that genAI systems can be helpful in learning to process and express concerns with those that we respect...so that we're learning to be authentic and helping others expand their world models. In the past, I would send some long emails to my academic peers to see if the ideas resonated with anyone...Now I post comments on substack as a way to explore resonance. When I write substack posts, some are done in collaboration with ChatGPT as I'm trying to weave complex ideas for broad audiences and want to make sure I minimize my personal bias and express the ideas in non-triggering ways. AI tools will continue to improve our ability to communicate complex ideas using powerful examples, just as you have done here. From a modeling and simulation perspective, it's important to learn how to question the results of AI systems and that can result in expanding our world models so that we come to understand we are deeply connected at the level of energy vibrations and that Kindness is the key for learning to improve our relational and transactional communications with others. I also play with using genAI for creating visuals to express the ideas that the universe is fractal energy patterns unfolding and that we can each learn to resonate with these energy fields so we can find ways to be orchestrators of attuned signals that align with sustainable futures for living systems. Keep up the good work!
Thanks Karen - and a great example of how considered use of AI with correspondence can be helpful - it's not all risk!
Really appreciate this post. Like most others reading this, I imagine, I enjoy correspondence and look at it as opportunity for authenticity and rapport—so I write all of mine with no AI.
However, I also try to set aside my own bias as an English teacher and confident writer, as the downsides of a poorly-written, authentic email are heightened when writing isn't a strength for you.
That is where I think this post is most applicable: for those who have to weigh real downsides on either side. Cool approach/lens—much appreciated!
Thanks for this Marcus - one thing I struggled with when writing this was that I know people who genuinely use AI to help with emails as English is a second language to them, or they feel unsure or anxious about their writing ability. Here I do think the tradeoffs between polish and authenticity are interesting -- and not always intuitive.
This is (one of many) reasons I don't use AI to write emails. But, more importantly, I find long emails don't generally get read. If there is something complex that needs to be communicated, I generally go the old fashion way - talk to them in person. Though it's probably what it's used for the least, most of us carry cell phones ...
I write my own emails because relationships matter to me. (And I like writing.)
The only way I've so far figured to message this in-medium (rather than through meta commentary) is to keep email messages so brief that it would be *at least as much* effort to write the prompt, edit, and cut-and-paste.
At that point it becomes credible that I wrote the message myself because it was, in fact, the incentivized option! Of course, this limits the domain of discourse... but I suppose I prefer to use email for shorter communications and, say, a proper briefing note, memo, or position paper, for anything longer.
I'm the same -- it's also interesting that short, personal and occasionally flawed are becoming the hallmarks of genuine human communication!