ChatGPT: Generating a Role for AI in the Workplace

Great events create great brands, and it takes a village to put on an event that engages, excites and connects audiences to your brand. And we're that village. I'm Alyssa. I'm Paulina. And I'm Rachel. And you're listening to Great Events, the podcast for all people interested in events and marketing. 

Paulina: Hi everyone. What's going on in the wild, wild world of events? My name is Paulina and I am here to welcome you to the Great Events podcast. This week we're going be chatting about all things machine learning. Most notably ChatGPT, and I am joined today by one of our recurring guests, one of our favorite recurring guests, Felicia Asiedu. She is also a fellow Cventer. Felicia, say hi.

Felicia: Hi, I am always so glad to be back on the podcast. For those of you who don't remember me, which is very odd if you don't remember me, I look after the marketing team in Europe. I speak at lots of events. We run Cvent CONNECT Europe over here, which is the counterpart to Cvent CONNECT US, which Paulina is so deep in the weeds of at the moment. But yeah, love all things tech. I've been in tech for 15, I don't even know, 18 years now. So it's been a very, very long time. So I love this conversation about AI and ChatGPT. Awesome.

Paulina: Awesome. Yeah. And so I figured for today's conversation, you know, we give our unique experiences with how this trend now, call it phenomenon is impacting, you know, our particular roles, the meetings and events industry. We'll contextualize it with maybe some best practices for all of you listening in. And then, you know, we'll talk through some of the things we should be cautious about and what is coming down the line. So with that, I'd love to hear, Felicia, your first guess interaction with using an AI tool like ChatGPT and what were your initial reactions to it?

Felicia: Well, I’ve got to say, if I'm being really honest, I was scared of ChatGPT and not because of the reason that some people would think with, you know, it's robots or AI taking over the world, just like those films that we used to see. But actually because it was a piece of technology that I didn't really understand and I didn't want to admit that I didn't understand what everybody was talking about, so I kind of played along with it and I was like, oh yeah, yeah. ChatGPT. Yeah, I know that.

Paulina: And I completely forgot for people who maybe are tuning in right now, having that very moment of their own, ChatGPT. Okay. If this is the first time you're hearing of this let us help you. Essentially it's a general purpose chat bot, right? So it uses artificial intelligence to generate text after a user enters a prompt. Essentially, the chatbot uses a language model that produces human-like or emotive text, and it's really all about, you know, giving a really strategic, comprehensive prompt for the output to be reflective of something that you're looking for. So basically you're chatting with a bot, you're putting in a question like, help me write the title of an essay on how to plan events, right. It will populate some suggested titles for that essay. Sorry to interrupt but we had to acknowledge that that group of people tuning in who might be in the shoes of what you're talking about

Felicia: Exactly, and I wish I had somebody by my side telling me that at the time, but I must say, even when I did get that knowledge that you've just given, and then I went on to the, you know, the generic website and I saw that single line of like, okay, hi. And I was like, so what do I say now? You know, hi. So I did some Googling and you know, listens to things like this and it said, okay, here's what you've got to do. You've got to say something like you just said, I need a 500 word blog about engaging attendees with header tags, H one, H two; it has to have this tone, emotive tone to just bring people along; it has to include these keywords. So once you start to give it that specific, like you just said, you know specific instructions, I pressed go and what it spat out blew my mind. I was like, this thing literally just did my job for me. Am I made redundant now? But honestly it was such a game changer because I realized I could just do things a bit quicker. I must say, we're going to talk about this a bit later, but I didn't, I'm glad I already knew I wanted it to sound a bit more like me. So went back in, made changes, spoke to the bot again, said, could you add this? Could you change that? But it just makes life quicker and I think that's what we're all aiming for a little bit, especially in these times where you're trying to be more efficient and save money and time.

Paulina: Yeah, that is completely the truth. Yeah. I think, you know, I think you're not alone when you felt like, oh, this robot, this chat bot could potentially do a segment of my job or my job. You know, seeing how quickly this, technology has progressed. But I think to your point the real goal here is it gets us 80% of the way there. Right, and it's about that 20% where you are critical because you're the subject matter expert and there is this tone of efficiency because you know, there's all of this, so many of us in the event professional role, event marketing role, we're charged with developing content in some capacity, and half the battle is just getting started, right? I don't want to call it writer's block, but you know, we're all doing so many things every single day. You know, you finally get to that one task on your list that's like, okay, now write a quick blog, or what is the attendee going to get out of this session? Right? And sometimes you just need that little kickstart, and this is really one of the most amazing tools to help call it that writer's block moment, if you will. I feel like, you know, I mentioned at the beginning of this, we really wanna contextualize conversations specifically for those of us in the meetings and events industry because there is a lot, I mean significant opportunity in finding, planning efficiencies, marketing efficiencies, customer engagement tactics that I think would really benefit all of us. I'm going kick it to you first, Felicia, to talk through some areas that you see there being enormous impact in the event marketer perspective.

Felicia: Yeah, I think it goes all the way from a basic level all the way up to a complex level. I don't think we're at the complex level yet, but I'll start with the basic level. So like you just saying, spitting out information that it's like, I need session titles, and then I need the description. I mean that if we look at Cvent Connect US, how many sessions would you say are, you are running at Cvent Connect us this year? Roughly.

Paulina: Oh, 75.

Felicia: 75 sessions. So that takes either a lot of brain power to just be sitting there, coming up with 75 titles and 75 descriptions. Or you and your content creators just move it along a bit quicker and then they amend. So I think it's those bulk tasks like that, or all of the notifications that are going to go out to people during that conference. Breakfast is ready. You know, like all of that. It's just so much that you could probably just remove and stick in to ChatGPT to get you started. I think on the more complex side of things, I mean, I've heard some outlandish things about people putting all of their data, and I'm not going to tell you this is all ChatGPT, this is AI in general, but people putting their data of attendees into systems that analyze the attendees and say to the person running the event, your attendees are a bit like this. They seem to like this. They seem to want that. They seem to want to hear this. And then people are taking that information and amending everything from keynotes all the way through to content, to timings and understanding more about their audiences prior to the event. I think that's fantastic. If I could do that, that is complex, but I would all day. We've been trying to personalize for years, so that would be amazing. But I think for now, it's the simple tasks that we're probably going to focus on today.

Paulina: Yeah. I mean, and simple tasks that actually end up taking a significant amount of time. Right. You know, as you look at, you mentioned push notifications, we think about push notifications really should send more than six per day across your conference, right? And so half of those are likely calls to action for behavior, right? You want people to get to breakfast. You want people to go to sessions. Oh, the awards gala is starting. The other half are probably in some capacity targeted towards sponsorships, right? Or, you're promoting something that isn't necessarily a behavior, right? So, you know, don't forget to take the session survey. That is actually going to be more of a behavior tactic. But as you think about it, if you could get 50% of the way there by utilizing a tool like this that is a significant lift off of one of the tasks that sometimes take significant time in coming up with thoughtful, but also within the context of the limitations of text messages or push notifications. It can only be 140 characters and sometimes 140 characters is half of the sponsor's title. So I think there's a lot of support there that can be helpful.

Felicia: Just to add to that point, I know, and I know it's not something we should freely admit, but that's probably one of those tasks as well that gets left to the last minute because it's not a priority. And so we're thinking about all the other things and we're like, oh no, someone needs to write those push notifications. So time saver, but also a little bit of a rescue moment where you have left it to the last minute and you need something to just come along and be like, don't worry, I've got you.

Paulina: Speaking of rescue moment for those who do showcases, trade shows, that have some kind of exhibitor component. We all know that there are sponsors and exhibitors who just forget to upload their logo or send the right format of their logo, right? If there is a machine learning tool, an AI tool that can, if you say I need the, the Cvent logo in a vector format, with a white background. I mean, can you imagine being able to just quickly aggregate all of the missing logos and being able to say, okay, I confidently know that this was pulled at least from a reputable source, right? Or, hopefully it's from a reputable source. You're able to either take that logo and send it to the point of contact and say, can you confirm this is the right logo? Or let's say that you're just being ghosted, at least you have something and they're not getting a blank logo or representation, whether it's in the mobile app or onsite. I mean, so many lifesaving quote unquote lifesaving tactics here. But, I think that one would be really interesting. I think on the more sort of complex side of things, we have a very comprehensive, if you will, staffing process for all of our events. We utilize a lot of internal staff to support our onsite logistics for our conferences and in some cases people are being pulled into roles that are really out of scope for, you know, what they do on the day-to-day basis. And what we do as a meetings and events team, we create a playbook. I mean, this playbook is, you know, six pages and longer in some cases - full of information. It's the entire conference information. It's, oh, are you going to be a session scanner? Here's the role of a session scanner. Oh, are you going to be moderating a session? Here are all the responsibilities for moderating a session. And each year we iterate off of it, right? The tech gets a little bit better. Maybe you're in a new location, the dates are different, right? Taking a first swag at updating a staffing playbook or quite literally any kind of program playbook would be a huge lift off of a team member's shoulders, in terms of updating sort of those baseline details. But then if we're able to take it a step further and say, what are the main responsibilities of a session moderator moderating a 35 minute session with polling and Q&A? Wouldn't that be amazing if all of that would be provided as like a one pager? I just keep thinking about like, sort of like the minutiae roles and responsibilities or tasks that us event profs have and how we could really take it off of our plate and see how tools like this can help us.

Felicia: I love that so much. And even with what you're saying, like I'm thinking about the fact that you yourself, become more creative when you think about utilizing tools like AI. You're like, oh, actually I could do this. Because even as you were saying that, I was like, I know in other walks of life, I'm a governor at school and they ask us every year to do a skills audit. So they say, oh, what do you think you're skilled in? So imagine there was something that the staff of Cvent could say, I think I'm already skilled at this, I think I’m really really skilled at that, and they put that in and the bot says, well, we need this job to be done at the event because this work is really skilled. You know, it's something, it just does matching for your staff with your events. I mean the possibilities are endless at this point.

Paulina: Very, very much so. Yeah, I'm trying to think of any additional ones. We actually met as a team, about a week ago or so, and in just 20 minutes listed out all of the ideas that kind of came to us where AI could be helpful. I'm just going to list a couple additional ones off here just so you can think about it. We already talked about smart staffing, UTM tracking, automated SMS workflow tied to that push notification, any kind of messaging elements. Let's see here. Recommended integrations or smart planner alerts. So when someone checks in, there's an automated message that's sent to them. What are some of those messages that you'd want to send to those VIPs? You know, checking in, speaker headshot and biosearch generation, very similar to that exhibitor booth logo that I was talking about. Media asset creation. Let's see here. We had predictive lead flow, what's the next best action? I mean, hopefully you have a demand gen team that can help inform this, but let's say you're a small team and you're looking for best practices on how to navigate lead flow. I mean, this tool can certainly benefit. There's tons and in fact, maybe I'll actually see if we can upload this list, if you all are interested in checking it out. It's just something that the team threw together. We're obviously going have, I think, multiple brainstorm sessions around this topic as it continues to evolve. But Felicia, let's also talk about some of the areas we need to be sensitive to, just given, there's a huge scope of support that this tool offers, but there's also areas that we need to act cautiously. Can you give just a couple examples of what I mean by that?

Felicia: Yeah, absolutely. I think, you know, I'll start with the one that is easier to kind of think about, and that's plagiarism. I think if anyone wrote any essays at school and you know, your teacher said be careful about plagiarism at that time, even at that time, they had systems, if you're as young as me, where they could run your essay through a system and that essay would check the world, check the internet and say, has any part of this essay come from somewhere? Think of ChatGPT and similar things as the other way around. Obviously you are not feeding it to run and figure out did you plagiarize, but what is actually happening is as you put your request in, it's going to the internet and pulling from there and then giving it to you. So in fact, it's kind of working in reverse. So you have to be really mindful of the fact that if you are writing or asking it to write any long form content, that you are going to go and publish on your websites and all that kind of thing, you need to kind of look at the text that comes back and make edits. Strong edits. Not one or two words, but really strong edits. It needs to sound like you and it should anyway cause it's probably not going to be on brand. You know, there's probably people that want to collaborate and make sure that it sounds like your event or whatever your theme is. So ChatGPT is not going to know that. So you have some work to do. I would just caution and advise that you can't just use what you get. I would say that's the first thing.

Paulina: I think those are really, really great points. And I think the biggest thing to take away from our conversation today is that this is a tool to assist. It's not meant to replace, it's not meant to say that it does the job for you. So as you think about all of the prompts and the content and the detail that you put into it, be really specific. Be comprehensive. I think this tool has an opportunity to help us potentially eliminate bias in terms of how we write copy and content. But it also, if you don't have a really strategic prompt, can help perpetuate it too. So there are certain things that I think are really interesting and how this tool will continue to evolve and how it can impact the way content is written and thus received. But I feel kind of alluding to all of the future of what this tool can do, you have a very cool use case scenario of, not use case scenario, don’t listen to me. A unique scenario of what the future can look like specific to facial recognition. What, just tell us all a little bit about that.

Felicia: I will do so. There's two things that I came across recently. One we signed a great partner recently where they have facial recognition technology. They're called YEG, and they're based in Italy. They already had this technology, but what they were really excited to do was to integrate it with Cvent technology because we can track registration, arrival, where are they going? We've got beacon tracking and are they in this space or this space? And then they said, if we put our facial recognition technology with your tracking technology, we don't only know when they came, where they are, but we know how they are, are they happy in that space, at that time, and when were they happiest? And I was like, whoa, this is like a whole new world. But they want to really give that 360 degree view of every individual attendee that comes. And that though we're talking future, is today. Like they're literally doing that right now, integrating. So I can't wait to see if we can kind of roll that out further than Italy and see if we can get to use that app, maybe connect Europe this year. But also I went to the MPI European conference and the keynote speaker said that he had run all of our data through a system, an AI system, to figure out who we were and what we were interested in. And I was like, wow, okay. And he said, so he changed his keynote. He was going to talk more about say, direction A, and he said, but this audience is far more interested in direction B, so I'm going to tailor my keynote before I get there to direction B so that it lands with the audience. Cause I know you're interested in this and this and this. And again, I think I felt a little bit if I felt a bit invaded, like how do you know me before we even shook hands? But, amazing because I was so fascinated by what he had to say, and it's probably because he did his research. So I think these areas of personalization are what we are really going to use as time goes forward. But I was also thinking, what if it goes even further? And just like our smartwatches, what if somehow we can use AI to track how we got to the events? Was it by car, plane, train, or bus? And how much carbon did we expel by, or expend by doing that? And then could that be given to you, Paulina, and say, well, here's all the people that came to the event and here's how they got there and this is what's happened. Or by that vegan buffet that you put out thinking you were going to do well by, you know, the world, there was the facial recognition technology saying these people are not happy.

Paulina: A lot of frowns. A lot of frowns.

Felicia: or vice versa. Maybe they're just nonchalant and they don't really care. And actually you can get away with doing vegan menus more because they were quite fine. You know? Or, you are happy eating that food. So there's so much that can happen.

Paulina: It's so interesting to hear you talk about things that we typically capture in registration, but you know, as a marketer, as an event professional, what is one of the number one goals when someone's going through registration? That they completed. Right? And so if this is able to compliment registration questions in a way that you are really spearheading the registration process and making it very simple, easy, quick hit questions, and we're able to collect this data in another complimentary way, I think that might be a beautiful compromise of what's to come. Because every marketer's first goal is, get that registration. Don't ever have those, what do we call it? The incomplete registrations, abandoned registrations, and so maybe AI is the resource for tapping into those abandoned registrations, instead of having sales utilize really valuable time on the phone and asking them to follow up on it. AI is able to navigate the rationale or the reasons why. So I feel like so much to unpack here. There's no way we'll ever cover all of the goodness on this topic, especially at the clip of it, how it's evolving. Maybe we'll do a part two, part three throughout the rest of the year and come back to this. Maybe we'll do a readout on how, you know, we've utilized machine learning for our own event planning purposes, for our US event, but definitely for our Europe event. But yeah, I think we talked a lot about great use cases for integrating all of these machine learning, opportunities in our day-to-day. We hope you enjoyed our chat today and found some inspiration, some takeaways. As always, if you have questions or topics or people that you would like to join us on the podcast, please send us a DM. Send us a note on LinkedIn. You can always email greatevents@cvent.com. I'm Paulina. Thank you so much, Felicia, for joining today. Loved the quick conversation and look forward to future conversations on this topic. Thanks all.

Felicia: Thank you.