Using AI to Improve Your Communication (ep.145)

Do you use AI to boost your communication effectiveness? Andrea shares her evolving perspective on AI, including whether using AI is “cheating,” practical ways to use AI to improve communication, and offers three “Dos” and three “Do Nots” for effective AI usage. Do you have more to add to the conversation? Let’s talk!

Resources

• Forbes October 2023 article on AI
• Talk About Talk ep.15: AI with Avi Goldfarb
• Talk About Talk ep. 129: Power with Tiziana Casciaro

Connect with Andrea & Talk About Talk:

• Website: TalkAboutTalk.com
• Communication Coaching Newsletter:  https://talkabouttalk.com/blog/#newsletter-signup
• LinkedIn: Andrea and TalkAboutTalk
• Youtube Channel: @talkabouttalkyoutube
• Talk About Talk Podcast Archive
  

  ---

   

Transcript

Let me start this episode by staying until recently, I didn’t think I had a lot to say about using AI. Four and a half years ago, I interviewed Professor Avi Goldfarb for a Talk About Talk podcast episode on AI. Avi is one of my past colleagues at the university of Toronto, and the author of two AI-focused books, Prediction Machines and Power & Prediction. This was before we’d heard of ChatGPT, and before everything really took off. Obviously a lot has changed since then.

Then, in the past month or so, I had a couple of experiences that made me realize, yes, in fact, I do have a thing or two to say about AI.

Recently I was collaborating with an organization to lead a series of workshops for them. After our first meeting to plan the workshops, they sent me an e-mail asking for a title for the workshop series. I thought about it for a few minutes and nothing inspiring came to mind. So I opened ChatGPT, and I asked it to generate ten possible titles for the Workshop series. I told ChatGPT who the audience was and what the topics would be.

The list that ChatGPT generated was boring. Not helpful. So I asked the same question again, but this time I clarified that I was looking for CLEVER titles. This new list had a lot of potential. None of the 10 suggestions were bang on, but it was incredibly helpful. This list inspired me and helped me generate some better ideas.

After I forwarded 3 potential workshop series titles to my client, I quickly got a response  saying that the titles were fantastic.  Frankly, I agreed.

The next day I shared this story with a girlfriend of mine. She’s a lawyer and she works full time in a big organization. So her response surprised me. She said, Andrea! That’s cheating!

What? 

You’re not allowed to do that.! That’s cheating!

Wait! I asked. You think using ChatGPT to help me come up with a title is cheating?

Yes! She answered.

Oh boy. I had a lot to say about that, let me tell you. So that was one experience.

A few days after this conversation, I was at my sons high school, and I met his English teacher. He asked me how everything was going and if I had any questions for him. I know my kids would always wish that I’d keep my mouth shut, but that’s just not my style. So I asked him, what’s the policy of your high school students using AI for their homework and assignments in your English class?

UNFORTUNATELY, this high school English teacher’s response was exactly what you might guess.  And that disappointed me. He started giving me the speech about what AI is. OK. Then he started in on the potential and the downfalls of the technology. And then he said, “For now, the policy is that it’s not allowed.” 

I just looked at him with a completely neutral face and nodded.

It’s certainly not that I didn’t have a response. I had a lot going on in my mind that I wanted to say. As it turns out, I do have an opinion. 

So those two conversations are what inspired this Talk About Talk episode #145 Using AI to boost your communication.

---

Greetings and welcome to Talk About Talk episode 145: Using AI to Boost Your Communication. 

In this episode, you’re going to learn:

My thoughts on AI and cheating

How you might use AI

And some Dos and Don’ts when it comes to using AI to boost your communication. 

My hope is two things: 

1. I hope you’ll be inspired to use AI more
2. I hope you’ll learn something new to try

Before we go any further though, I just want to say that I in no way claim to be an AI expert. Rather, I’ve used AI and done some reading on it, and as I said, I have a few things to say. if you’re a tech expert, this episode is probably NOT for you. I’m certainly not a tech expert. 

I acknowledge I have a LOT to learn! SO please connect with me on LinkedIn and join the conversation. Tell me how YOU’RE using AI to boost your communication.  I’d love to hear.  Bring it on. 

OK – I better introduce myself. In case we haven’t met, my name is Dr. Andrea Wojnicki and I’m YOUR executive communication coach. Please call me Andrea!  I’m the founder of Talk About Talk, where I coach communication skills to ambitious executives through 1:1 coaching, bootcamps, workshops and keynote speaking. My objective is to help you improve your clarity and confidence, so you have more credibility. When you have credibility, you can make real impact. And that’s when you’ll get noticed and your career will take off!  Sound good?  

If this resonates with you, then I also encourage you to check out the Talk about talk.com website. There are many resources there to help you out. If you’re an individual executive, there’s information about online courses, private coaching and small group bootcamps. If you’re a leader or an HR manager looking to boost the communication skills of your team, there’s also information about corporate workshops and keynote speeches.  And there are plenty of free resources too: like the archive of this bi-weekly podcast, AND, I really hope you’ll sign up for the Talk About Talk email newsletter. That newsletter is your chance to get communication tips and coaching from me every week.  I also  hope you’ll connect with me on LinkedIn and maybe send me a message and let me know what you think about this episode.

I’ve experimented quite a bit with ChatGPT and some AI-based image generation software. I’ve also done some research (listened to podcasts, done some reading), and I’ve done a lot of thinking about this topic. And as I said, these experiences that I shared at the beginning of this episode, (1) My conversation with my friend about whether using ChatGPT is cheating, and (2) the conversation with my sons high school English teacher — These conversations made me realize I do have something to say about AI. And I’m hoping that I can encourage you to think about AI and use it to boost the quality of your communication.

In some of the reading that I did about AI, I learned some interesting trends and statistics that helped me put things in perspective. In an October 2023 article from Forbes Magazine, I read that 58% of executives say they use generative AI regularly. Does that sound hi or lo to you? I thought it sounded about right.  58% are using it.

The same article said that 81% of companies have generative AI teams. That seemed high to me. Also surprising to me, is the fact that SMALLER companies are more likely to leverage generative AI, then larger companies, primarily due to perceived trust and accuracy issues. 

This might explain why my girlfriend accused me of cheating. She works at a larger company. Plus she’s a lawyer!

So here’s a big question. Is using AI cheating?

What do you think? Yes or no? Is using AI cheating? If you ask me this question, I will not answer you with a definitive yes or a definitive no. I will say that AI itself is a tool. So AI cant cheat. It’s the person using AI who can cheat. 

This reminds me of a conversation that I had with Professor Tiziana Casciaro about the term “power.” I interviewed Tiziana for TAT episode #129. I’ll leave a link to that episode in the shownotes.  Tiziana said that power itself is benign. It’s what you do with power that makes it good or bad. Power is a tool. Kind of like a hammer. You can use a hammer to build a beautiful home. Or, you could use a hammer to seriously hurt someone. It’s what you do with it.

The same holds for AI. AI is a tool that can definitely, most certainly, be used for good – and for bad. Personally, I am optimistic about it. I do believe that it’s going to change the world in profound ways, and I’m very glad that there are some smart people working on regulations and policies to make sure that this tool doesn’t become a tool of destruction.

But I haven’t actually answered the question yet, have I? Is using AI cheating? Well, let’s go back 30, 40, or 50 years ago.  

Imagine yourself in a senior high school or university setting. What were the big topics of debate? Well, how about encyclopedias and calculators.  Some of you may have never seen an encyclopedia.  But I remember in grade 8 or so, writing an essay for history class and using an encyclopedia to help me.  I was very very careful not to plagiarize.  We were taught to read, then think, then look away and write.  No plagiarizing,  No cheating. 

The encyclopedia was the resource, or the tool. The onus was on the students to not cheat.  Do you see the parallel?

An even bigger deal in schools back then were calculators  When I was in high school and university, there was a big uproar about whether students should be allowed to use calculators. I remember watching about 5 of my classmates in my operations management final exam get walked out of the gymnasium for cheating with their calculator.  Yikes. 

Nowadays my 15 year old daughter has a mighty powerful calculator in her backpack. She uses it for assignments AND in exams. It’s a tool.  It’s definitely not cheating.

Something that I noticed when I was doing the reading about AI that I guess is kind of obvious, but I just wanted to point this out is that for high schools and universities, the conversation about AI is really focusing on catching plagiarism. The fundamental assumption here it seems is that students who use AI to generate essays, for example, are at fault for plagiarizing. We blame the user.

Interestingly, when you shift the lens to business, it’s less about plagiarism specifically and more about copyright ownership. It’s less focused on blaming the user, the business person or the organization who leverages AI. It’s more focused on blaming the technology itself. What sources is the AI referencing? And what are the copyright obligations of the AI technology to the rightful owner of the intellectual property. 

Does that make sense?

 I find this fascinating. We’re ultimately talking about the same technology being used in similar ways. But when the context is essay writing for high school and university students, the main focus and blame, is on the student for plagiarizing. 

In a business context, where an essay or a presentation may similarly be generated through AI the focused and blame is on the technology for not attributing the copyright properly in terms of the original owner of the intellectual property. Fascinating.

So let me answer the question.  Is using AI by-definition cheating? No.  AI is a tool. It is possible to cheat with this tool though, if you generate content using AI and claim it as your own.

So was I cheating when I used AI to help me come up with the titles for the workshop series I was working on? No.  I asked ChatGPT to come up with some clever titles, and I used those ten ideas to come up with a few titles that worked very well.

You could say that I the titles were AI assisted.  This is another insight that kept reading about you probably already know but I think it bears mentioning here.  It’s the distinction between AI generated vs AI assisted. 

I have an example here that I think beautifully helps illustrate the difference between AI generated and AI assisted.  When I interviewed professor Avi Goldfarb about AI, he talked about how the professors he admires are allowing students to use AI as a tool in their essay writing.  He talked about how, for example, a professor would assign an essay and share with the students what an AI generated response to the assignment would look like.  Then the prof would say “Something similar to this would probably get you a C. So the baseline has gotten tougher.  I expect you to use AI and your own knowledge and creativity to come up with something significantly better than this.”

The essay that the professor generated using AI is exactly that – AI generated.  The versions the students would turn in would be AI assisted. Ill get to some of the ways the students could use AI in a moment.

Avi’s example of how professors and teachers might use AI in the classroom is the answer I was hoping my son’s English teacher would have to my question about AI in his class.  Maybe it’ll just take some time.

In business, some folks may start with AI generated copy, then edit it.  That’s still called AI generated.  Other folks, including myself, create content, then use AI to edit and improve it.  That’s AI assisted. 

By the way, if I ever use AI generated content, I will let you know!  That goes for copy and images. For this episode, I used ChatGPT to assist me with the description and I used Lensa to create the feature image 

Now, I’m going to share with you some ideas for how you can use AI to help you communicate in business.  Then I’m going to conclude with my 3 favorite ways to use AI, 3 DOs and 3 DO NOTs when it comes to using AI.

THREE FAVOURITE THINGS TO USE IT FOR: beyond spell check and grammar check and generating basic emails, contracts, images, and transcripts.

1. BRAINSTORMING IDEAS
   • I use AI to make sure I’m not missing anything when I’m listing suggestions for you.  Like the episode I released a long time ago on how to dress at work.  Episode #xxx. I wrote the outline for the episode, then I asked ChatGPT to list ten suggestions or rules for how people should dress at work. I used the list to make sure I wasn’t missing anything.
   • So you can use AI to make sure your suggestions are complete.
   • Another way I’ve used AI is to suggest relevant metaphors an analogies. This can be incredibly helpful in elevating your communication.  Sometimes the best way to describe a new concept or an idea as similar to something that your audience already knows.  Like how AI is a tool that is similar to encyclopedias and calculators, in some ways.
   • So that’s the first way to use AI.  For brainstorming.
2. REFERENCES –
   • Much of the advice I share in coaching and workshops is grounded in academic research. I often use AI to help me quickly identify research that either backs up my advice or that supplements it.  Chances are you are not in academia and you don’t need to cite everything.  But you can still use AI to back up your recommendations at work with objective references.
3. SUMMARIZING – 
   • I use AI to help me write podcast episode descriptions. These always need a heave dose of editing, but it’s a good start.  You might use AI to help you with summarizing a presentation or a white paper.  Or even for data analysis.  That’s summarizing.

 

So those are my 3 suggestions for how you can use AI. Brainstorming, References and Summarizing. That’s a great start!

Oh – this is not one of my 3 things, but I do have an idea that I’ve shared with a few clients that I want to share with you.  It’s this. Automatically generating transcripts from the audio recording of your presentations or meeting dialogue and then analyzing the transcript as a means to improve your communication. Here’s an easy hack.  Record the audio and then search the transcript for your crutch or filler words.  How many times did you say “like” or “um” or whatever your crutch word is? Similarly, if you’re working on avoiding upspeak, you can search the transcript for question marks, when you were asking a question,  That’s the definition of upspeak.  Helpful, right? (CATHY, THAT’S A SOCIAL MEDIA POST!)

There are also are some new platforms and software out there that will take this kind of insight to your communication skills to a whole new level.  I’ve used a platform called YOODLI to analyze some of my clients formal presentations. You upload a video and it assesses your tone, your cadence, your engagement, your crutch words and more. The data generated by Yoodli is supplemented by my suggestions.  Its kind of cool,  Let me know if you want to give it a try!

OK, moving on.

Now I have 3 Dos and 3 DONTs for you. Starting with

Three DOs:

• (1)Consider the multitude of ways that AI can enhance your communication.  Consider the media.  Of course AI can help you with words.  There’s also audio, images, videos, and more.  So explore various media.  Also consider your various tasks and the folks you work with.  Ask them if they’re using AI, and if they’re not, encourage them to try it out.  
• (2) Focus on getting good at writing prompts. Practice. If you get really good you might even become a Prompt Engineer. Yes, that’s a real job. You see, AI isn’t just taking our human jobs away. It’s also creating new jobs. Anyway – my prompt writing is slowly improving. I always assume it will take me several cracks at a Q before I get something back that’s helpful. One thing that I’ve found really effective in writing prompts is to add the tone or to say “in the voice of”  This is what I did when I was genereating the titles for the workshop series.  When I added that I wanted “clever” titles, they suddenly improved.  So DO practice writing prompts
• (3) This Is a new one for me: Encourage ChatGPT (or whatever software you’re using) to ask you Qs about what you’ve asked it to do. Like when you’re assigning a new project to your direct report .  You might ask them at the end to summarize what they’re going to do  You could ask your AI to do the same. “Please summarize in different words what it is that I’m asking.” Then, what other questions do you have for me about this request?”  This is a strategy that will help you accelerate your skill in writing prompts.

And last, 3 DO NOTS

By the way, I almost always find the DO NOTs to be the most helpful.  Have you ever noticed that? This is an aside, but years ago when I was learning how to teach, I asked one of the top professors at Harvard Business School for any advice she had for being a better teacher in the classroom. The first thing she said was, ask lots of people this question, andrea.  But focus more on what they tell you to avoid or what NOT to do. What you SHOULD do is probably personal,  What you should NOT do is more universal.  TRUE.

SO what should we NOT do with AI?

(1) Don’t assume AI is correct. It can be very wrong! A while ago I created an experiment that ended up illustrating how ChatGPT can be very wrong. I asked ChatGPT to summarize what the personal brand is of Executive Communication Coach Andrea Wojnicki, based on her LinkedIn profile. Based on the time lag of the data that’s been uploaded to ChatGPT, I assumed that the summary might be slightly out of date. But I was surprised to see that the summary generated by ChatGPT was categorically wrong. It included facts about my background that were not even hinted at in my LinkedIn profile. For example, I remember it said that I had previously worked as an actor, which helped me in my role as an Executive Communication coach. Nope. That ain’t true. So always, always, always Fact Check what AI is generating for you. Then, verify these “facts” with other sources.

(2) DO not ignore AI. Many AI tools are free right now and relatively easy to use. But that’s not the only reason to not avoid AI.  When I was doing some research for this episode, I kept adding again and again and again that the most productive and employable folks ar using it now. SO jump in. Here’s an insight that might inspire you: “Instead of losing your job to AI, consider that you might be more likely to lose your job to someone who uses AI better than you do.”  So do NOT ignore AI.

(3) LAST: Tell it what to do, not what to NOT do. As in when you’re prompting ChatGPT, for example, avoid negativity. Instead of saying “without sounding arrogant”, try prompting “in a modest voice.”  Or instead of saying “without being obvious or boring” try prompting “creative” or “clever.”  You get  the idea.

So the 3 Do Nots:

1. Do not assume AI is correct
2. DO not ignore AI.
3. Do not tell AI what to void or what not to do.

And that is it for this episode!

I’ve been thinking about doing an episode on this topic for a long time. I know many people know a lot more than I do about how to use AI.  But I also know I have something to say.  I hope I’ve inspired you in terms of your opinion about AI and cheating, in terms of the distinction between AI Generated and AI assisted, in terms of the many ways you might use AI at work, including brainstorming references and summarizing, and in terms of some of the Dos and Don’ts of using AI.  DO consider the many ways you can use AI. DO practice getting good at writing prompts and DO encourage generative AI to ask you Qs about what you’ve asked it to do. DO NOT Do not assume AI is correct, DO NOT ignore AI. And be positive. Do NOT tell AI what to avoid or what not to do.

I’m sure you have lots to add to this.  Pleas bring it on! I’ll be posting lots about this on social media and in my email newsletter. SO please connect with me on LinkedIn and share your thoughts! And please go to the TAT.com website and sign up for the newsletter. You can reply directly to the newsletter and we can talk more.  Bring it on.  Let’s Talk!

Thanks for listening.  And talk soon!