AI email etiquette. How much is too much AI?

Tough Things First

It is getting easier to instruct AI tools as a shortcut for sending out emails and social media posts, but when is it a misleading crutch? Ray Zinn discusses changing AI use and how to recognize the gray areas of using AI and calling it your own work. (Watch Video Podcast…)

---

Ray Zinn: Good morning there, Rob.

Rob Artigo: Well, this is a special edition of the Tough Things First Podcast. If you’re listening, you can check out toughthingsfirst.com, find this podcast, and right next to it there’s a link that says video. So you could click on it and watch it if that is your choice, or listen to it now and watch it later, whatever. While preparing this podcast, Ray, I logged into Microsoft Copilot, something I’ve been kind of experimenting with in a few different ways. And I asked it to write an email about this podcast, which is about having AI or ChatGPT write your letters, emails, or social media posts.

So I logged into Microsoft Copilot and I just said, “Can you write an email for me?” It said, “Yeah, what’s the topic?” And I told it what the topic was, and then what it said was… Here’s the email. Subject line is, “Join us for an insightful podcast on AI and email communication exclamation point.” And it says, “Hi recipient’s name, we are excited to invite you to our upcoming podcast episode where we dive into the fascinating world of AI-driven email communication. Our special guest, Ray Zinn, the longest serving CEO in Silicon Valley history, will be sharing his invaluable advice and insights.” And then it has the date and time and it says, “Put the link here.” And it says, “Look forward to having you with us. Best regards, your name.”

So what I’d really like to do is ask you just based on that, I mean, that was a very quick thing. If I didn’t want to craft it myself, seriously, Ray, it would’ve taken me just as long to write the email myself and do that same thing rather than having the AI do it. But let’s have your reaction to me going through the process of having something else create the email for me.

Ray Zinn: Well, for one thing, it says AI, so there’s the starting point. I mean it’s already doing kind of an advertisement for artificial intelligence. And it’s just interesting that AI really is a buzzword. We’ve had ways of doing emails that way before cut and paste. I’ve used cut and paste for, I don’t know, 50 years.

Rob Artigo: And people don’t think of that as being AI, but that’s basically what that is. It’s using an algorithm to work something out for you.

Ray Zinn: Yeah. So my point though is that I’ve pulled up old emails and just edited a few things and sent it back out. So if the topic is general in nature, like, “I’d like to invite you to my birthday or my anniversary or whatever.” That’s pretty straightforward because we have these group lists that we can call up and do a blast email blast as you would. Because this is an election year, I get probably 10 email blasts a day, and they’re all quote-quote AI crafted as you would. And I can tell it because they look bland. Okay.

Rob Artigo: Okay.

Ray Zinn: And so it doesn’t sound like anything that comes across as being personal to me as you would. The other day, I received a letter, actually a card, but written on it. It was handwritten by the governor of Montana who was thanking me for all the things I was doing in the community. That meant more to me than getting an email blast that just says, “Thank you for your support, blah, blah, blah.” And so handwritten, I mean, if I get a handwritten birthday card or a handwritten note from someone, that means more to me than if they just… Because I can tell if it’s crafted by AI or not AI, but by some generic email blast as you would. So can AI help? Yeah, sure, it can help.

It depends upon what you’re wanting to do, what your goal is. So if you just wanted to notify everybody that there’s going to be a meeting at such and such a time without having to sit there and type in that information, then you can do that with various, what they call abbreviated… You can type in, for example, M-A-T, MAT, meeting at time. You can have that just type in those three letters and it’ll craft, if you already have a predefined thing you want to send out. It’ll do that for you. And then you can just type in the time or the date. So this is really what it is.

As I’ve mentioned, AI has been around for a long time. I mean, it’s not… it’s just a buzzword referred to as AI. I remember Al Gore claiming he invented the internet, which was another misnomer as you would. The internet in a sense has been with us for many, many years and long before Al Gore did it. So what we do is we have these little buzzwords, these little things that we call smart this, smartphones, smart speakers, smart ovens, smart refrigerators, smart appliances. And they’re just nothing more than just a bunch of buttons that you push and it’ll then create whatever you want. For example, if you have a certain kind of washing machine load you want to do, you can have a predefined, you can program it as you would. It’s the same thing with your phone. You can program your phone to do different things or your computer.

And as I mentioned, this sort of thing has been around a very, very long time. I can remember back in 1978, 79 timeframe when IBM, what they call their… Oh shoot, I forgot the name of it, but it’s like a Skunk Works type thing. It wasn’t Skunk Works, but it’s had another name like Skunk Works, where they were playing around with doing a little personal computer. Now, IBM Corporate, they’ve been building computers since I was in college back in the 50s, but they were mainly for big corporations, and they were large expensive systems. They’ve taken up the size of a room. And so the Skunk Works Group down in Boca Raton, Florida, worked at IBM, wanted to come up with this little PC. And what’s interesting is that Heathkit made a PC, wasn’t called a PC, but it was a computer that you could put together, but that was a hobbyist thing. And all IBM did basically was take that S-100 bus, that Heathkit hobby thing, put it together with their own BIOS, and came out with a personal computer.

Now, granted it had a keyboard and a display and all that sort of thing, but it was nothing more than cobbling together stuff that was already there. And so when we talk about AI, you’re just cobbling stuff that’s always been there, making it a little more convenient and easy to use. But artificial means not real, fake as you would. And so fake intelligence, that doesn’t even sound good, does it?

Rob Artigo: [inaudible 00:09:04].

Ray Zinn: And again, if you’re just trying to knock out something, trying to speed up your day as you would, you can cobble together an email blast as you would by just having these predetermined emails that you’ve written over the past year or two and save them in your file and then have them named something, a certain name. You call it up and you just change the dates and times and so forth, or maybe a name or whatever, and you can do the same thing. So-

Rob Artigo: Let me just let you know that I actually thought of this topic when I was sales pitched by Adobe. And this isn’t a complaint about Adobe. Excuse me. I like their products, but what they asked me was, did I want to use their AI tool to write my emails? And what they suggested was that if I get a lot of documents, what this tool would do is it would scan the email for content and it would cr… I mean a document for content, and then it would craft an email about the content for you. And I said, “No, thank you. I can write my own emails.” What bothered me was the fact that they were interested in scanning my document and then crafting an email based on that. If I’m doing that, it probably means I’m cheating the system and not reading what’s in the document.

So if I don’t know what’s going on, and I put my stuff in the email based on what the AI suggests, and I send it out to everybody, other people will understand what the AI pulled out of the documents and its attitude or whatever you want to call it, toward what was in the document. But I won’t say what I believe is inside the documents. And so that kind of bothered me. What about the ethics Ray? Let’s wrap it up by asking about the ethics of it. If I don’t say this is AI, if it’s something that I created, but I don’t say it’s… I mean if it’s something I used AI to create, whether it’s an email or a long letter or something that’s typewritten, but I don’t really know what the content is because all I did was sort of do a cursory read after I had it create the message, am I behaving ethically? And am I really just setting myself up for failure down the road when somebody asks me, “Hey, you said in an email, blah, blah, blah.” And I don’t remember writing it because I didn’t, AI did?

Ray Zinn: Well. I think we’re giving AI more glory than it really needs.

Rob Artigo: Maybe.

Ray Zinn: When you do a cut and paste, if you’re using that term, again, you’re not creating something from scratch, you’re just going back to an old email and you’re pulling out some information and then doing a paste, being putting it back into the current one you’re doing. Is that ethical? And so whether it’s cut and paste, and that’s all it is. If you think about what AI is doing, they’re capturing tens of thousands of files and then creating something from you from those files. You can do the same thing just from your files, the documents that you keep. When you ask a computer or some other smart device, “What’s the date of Theodore Roosevelt’s birthday or whatever?” It just goes and searches millions of documents and brings back the information. It’ll also tell you the source of that information. It says, “According to Wikipedia or according to such and such and such source.” It’ll tell you what that source is.

So the ethical or the unethical part is that if you don’t say what the source of your information is, to some degree, that’s unethical because you’re pulling from somebody else’s information and using it as your own. So how authentic is that? Or how ethical is that? That depends upon what you’re trying to do. If you’re writing a original book and you’re copying somebody else’s work, then that’s not ethical. So when you copy somebody else’s, if they’re going to pull an email out of the universe and then say it was yours, that to some degree that’s unethical because it’s not original. So AI, if we know what… If something is specified as AI, then we can definitely say this is not original. So whether it be a video that looks like you, that’s not original. And so it depends upon what we’re trying to do, who we’re trying to fool as you would. So that’s the bottom line about what’s ethical and what’s not.

Rob Artigo: Yeah, and we talked at a previous podcast about what is copyrightable, and we know that portions of whatever your work is that you used specifically for… that you used AI to create instead of your own creative juices, if you will, then you’re not able to copyright that material. And that’s the reason why, because it could end up being somebody else’s material in the long run. And I think there’s nothing wrong with saying, “Hey, I used AI to create this.” And you could say that. You could just do a little disclosure, say, “I used AI to help me out with this because it was a long thing, but I read it and I agree with it.” So you could do that, and then that would kind of help you work it through.

Anyway, thank you, Ray. Great podcast, I’m glad we could do this on video. And I’d like to let the listeners know that you can join the conversation at toughthingsfirst.com. Your questions and your comments are always welcome. Follow Ray Zinn on Twitter, Facebook, I should say X, but I guess they still call it Twitter, Twitter, Facebook and LinkedIn. And of course, pick up Ray’s books, Tough Things First, and the Zen of Zinn Series 1, 2, and 3.

Ray Zinn: Thanks, Rob.