Justin: AI is the hottest topic in the world right now. That’s why I’ve gathered our directors here today to discuss this. Everyone’s using ChatGPT. It’s actually the most rapidly adopted technology ever, which is an interesting fun fact. I’m sure we’ve all been using it. So what do you think about it, David?
David: It’s awesome. It’s incredible how quickly it’s come along. Like for as long as we’ve had computers, there’s been the Turing Test, right? How to know if a computer is real or not. How does it interact with you? When I’m chatting with ChatGPT, I don’t know if I’m chatting with you or not. It’s not got the same sense of humor as you, of course you’re funnier. No offense.
Justin: I would like to think I have a little bit more depth than ChatGPT.
David: But just generally like you could easily be fooled into believing that you’re speaking to a person.
Justin: Except for when it says, “I am an AI generative chatbot and therefore I don’t have my own opinions and cannot comment on controversial…” Whenever it makes statements like that.
David: I see. Would you like to comment on anything controversial now?
Justin: No, because I’m actually a robot. I’m not a real human. I am ChatGPT personified. It’s true. It’s happened. Well, Max, what do you think of it?
Max: Well, it’s definitely changed the way we, let’s say, we gather information, I think, or it’s going to really impact the way we gather information. It’s able to just bring a lot of data together and to probably make some assumptions. Of course, there are still some issues. The data cannot be always trusted. We don’t really know the answer, where the answers come from, right? That’s one issue with most AI technology is that the output is not really measurable. There is no trace that tells us how it was generated. So it’s harder to fact check an AI. But yeah, nevertheless, it’s really impressive with the actual answer it can come up with, in every discipline. I tried not only in content writing but also in technical questions. Yeah.
Justin: It’s funny that you bring that up. It doesn’t always show its work. You know, as they said in math class when we were kids. It doesn’t cite any references either. So I’ve actually found that it’s kind of a liar sometimes. So we recently recorded our Lexicon Late Night episode talking about South Park. And I was just trying to think of catch phrases from Eric Cartman. I said, “Can you give me some catch phrases from Eric Cartman?” It generated a bunch of real ones. “Screw you guys, I’m going home.” “I’ll kick you square in the ____,” and so on. And then I said, “Can you give me more?” And then the rest of them were totally invented. It was like, “I’m not a liar, I’m just an entrepreneur.” And I was like, I’m pretty sure Cartman never said that. So like a lot of humans, ChatGPT would rather give you a nonsense answer or a lie than to say, I don’t know. I have yet to hear it say, I don’t know. I’ll test that out. Yeah, but I’ve found it the same as you. It’s fantastic as like, I think more as a writer, obviously you’re a tech guy. It’s a fantastic research assistant. It’s great at generating topics. It’s an excellent proofreader. If you’re stuck, it’s good at generating things, but it can’t really be creative on its own. It’s only as good as the inputs that you give it. Have you guys found that too?
Max: Well, there are literally courses on creating the right prompts, right, for ChatGPT and other AIs, and it’s really a skill to understand how to prompt AI, yeah. I think that it will become a skill of the future that you can put on your resume soon. Right. Right.
Justin: ChatGPT Whisperer.
David: Yeah. I’m impressed with the complexity of it. Like I can ask it quite complicated questions with weird clauses and stuff in there, but it seems to get it more than a human could.
Justin: Yeah, it understands you.
David: Yeah. Which is very rare. There should be a course on that.
Justin: Understanding David. That would actually be valuable for everybody who works at Lexicon. So I’m glad you found it as a useful assistant that gets you. What do you think this means for the world of marketing? Like, does this mean companies can only, can now they can just generate their own content and they don’t need people like us anymore?
David: To some degree. I think what ChatGPT does is take away a lot of the lower level labor that obviously you and I are writers by trade. This really helps us to write. So I think for many companies, they’ll be able to do basic stuff that they may have had to outsource before for a few thousand baht here and there. But, what this will ultimately mean, I believe, is that you’re probably going to have to have less staff. Every company will have to have less staff. So for us, we’ll be able to give away more of the monotonous tasks that we don’t want to do to robots, to computers, so that we can focus on really adding value to clients. Agencies are meant to be strategic partners, really bringing something different to the table, bringing creativity, bringing insight, and for us, helping a brand to develop its story, whether that it’s in visuals or in its tone of voice or in its execution, the whole 360 approach. The technology is going to make it easier for us, but also easier for clients. So I think that the trend over many years has been towards companies focusing on their core and outsourcing the rest. So everyone just wants to focus on doing their business. So companies probably won’t want to have writers in house when they can just have ChatGPT do the basic stuff. But that means there has to be somewhere else, an agency ideally, that houses the ChatGPT experts that can write, that can edit, that can add value. So I think it’s excellent news for agencies because it will remove a lot of the more monotonous tasks so that we can really do what we’re meant to do, which is help clients succeed.
Justin: You mentioned the Turing Test earlier. As far as writing something, I’m sure people are posting all over social media now with ChatGPT generated content. To be honest, I can’t tell. So that’s like as far as writing a caption, it certainly passes the Turing Test, except maybe I’ll know that it’s not a human because there’s no mistakes in it. So Max, we all know ChatGPT can write, but what about more tech implications? Obviously AI goes far beyond ChatGPT, but can AI code now?
Max: It can, yeah. And it’s actually quite impressive how detailed it can be in the solution that it can provide with a simple prompt. I was very impressed as well. It’s also really good at optimizing code so you can provide it with existing code and it can look for a way of optimizing it. It can, to some extent, write tests for your code. So it’s actually wonderful because all the boring stuff, it can do it very easily. Optimizing, testing, copywriting to some extent. But again everything is in the prompt. You have to give it the right prompt. And I think that’s where all the intelligence is defining the right solution and giving it the right prompt. And some prompt for coding can be 50 lines of describing what you actually want, right? So as David was saying, it kind of shifts the work into defining what you need and providing the right definition. And it’s actually the fun part is to work with clients to understand what they need and then to translate it into some kind of specification that can later become code. So although the coding part is still at its beginning, I think it’s going to change very quickly. It’s going to evolve very quickly. And knowing how to define an application and what it should be doing would be more important than writing the actual code soon enough, yeah.
Justin: I guess the same could be said about actual writing. There’s other tools as well that are pretty interesting. I’m not at all a visual person as anyone who’s seen me draw knows, but there’s some interesting things like Dall-E and Adobe Firefly Design AIs. Max, what do you, or David really, what do you think the future of a simple graphic designer will be when it comes to this kind of AI?
David: I’ll go first. Yeah. You can jump after that. So again, I think a lot of the more basic level graphic design should be able to be done by these tools, I would imagine. But it’s, again, it frees up talented designers to really focus on how do we develop branding, how do we push things further, how do we get to the next level? How can we make sure that brands, if everyone is using tools, then how do you stand out? That’s where the human part will come in. So those who are really talented, those who are really keen to perfect their art, they’ll be able to go to a higher level. But for the most basic level stuff, yeah, companies can do it by AI.
Justin: So that’s why creativity and strategy and storytelling will be at an absolute premium because anybody can generate the actual content. But what you need is for it to be cohesive. You need a big idea, you need a tone of voice. All of these things that we constantly tell our clients. Otherwise you’re going to be posting a bunch of content that looks good, but it’s not really telling your story effectively. Exactly.
David: You’re going to have a million ChatGPT generated captions and a million Firefly images. Are they going to sync up? Like marketing is still going to be a skill that you need to have. You still need to have all the stuff you described, right? What is the story your brand wants to tell? What are its values? Like, these are the things that you need expertise to guide that for you. So you need, I mean the role of a chief marketing officer is going to get even bigger and they’re going to have to have a few key people around them because no individual marketing person has everything. So I think for agencies, honestly, it’s very, very exciting where things go from here because we’ll all be getting more agile, we’ll be adding more value and we’ll be really helping clients to tell their story in more effective ways than ever before.
Justin: Yeah, totally. Focusing on the parts of our job that we actually find interesting and that we actually enjoy doing and that we’re better at, frankly. And then also proofreading, catching typos, kind of a thing of the past. Yeah, I think so. Well, I was going to kick it to Max. I think you’ve mentioned this before in your Ask the CTO session that we did here. I think we’re going to see the rise of the cyborgs, right? Because the AI on its own will be better than somebody who doesn’t know anything. The expert will probably be better than the AI, but the best of all will be a combination of the AI and the expert.
Max: Probably. Yeah. Well, I think AI, it will take a while until they have some kind of sensitivity and some kind of creativity, right? They can do what you tell them, but they can’t really direct what should be the direction. So it’s an amazing use for people that want to try new technologies. You won’t need to know how to use graphic design in order to create something. But as David was saying, if you want something that makes sense in the long term, you need a strategy, you need a vision, you need some kind of direction, and you need to be sensitive to the creativity side. So actually giving it input and direction.
Justin: Yeah, yeah, AI isn’t that sensitive.
David: I think it also just matches a trend, like I already mentioned, like focusing on the core for any industry, for marketing, for facilities management, for accounting, companies increasingly just want to focus on what they do and get rid of the rest. And I think we’ve already been seeing it for a long time in Max’s department with the web apps and the digitalization enabling tools that Max’s team has been developing, because I think that companies are going to want to be able to cut down staff to be as agile as possible. So I think things like web apps, they’re going to be even more important because they enable you to do so much that you’re going to have to be very competitive moving forward and being able to replace five people with a simple tool is, I think, every company is going to need that.
Max: And you need the data to be digital. That’s really the first step. Otherwise, all this AI revolution, you can’t really have access to it, right? So yeah, definitely having the data embedded in your day to day and step by step, adding AI to process, to pre-process this data, make it more valuable. And of course having someone that oversees the direction will be the near future, I think. Yeah.
Justin: I think we’ve kind of conveniently all made the case for why every single one of us will still be vital and we’ll still have jobs. We’ll be safe, for now anyway. But there is kind of a downside to this, right? Like you mentioned, replacing five people with one robot. Yes, it will enable companies to be more agile, to focus on their core, all of these great things. But what does this mean for the next generation? I’d imagine like recent graduates or people that are still in their fourth year of university and they’ll be entering the job market soon, this might make things tough for them.
David: Yeah, I think if you’re 15 years old now, amazing. Like you’re probably in a great place. You’re going to learn everything you need to know in high school and then you’re going to learn it at university and you’re going to come into the workforce with this amazing skill set, super optimized to succeed. You’ll be an expert in AI, you’ll be able to manipulate the technology and you’ll also have all these different economic options available that we don’t even know exist yet. We can’t even imagine yet. But if you’re coming into the workforce now in Thailand and you’re a junior and you’re 22 years old, you know, we’ve experienced that ourselves. Often a new graduate, they need a lot of time, right? They need a lot of mentoring, a lot of ‘Okay, go and do that again. Here are some notes for you.’ There’s a lot of our time involved in training those people up. And to be frank, ChatGPT can do that work much better. It’s equivalent to how young people today understand social media marketing.
Justin: It’s like having a thousand interns who never make a mistake.
David: So it’s going to be very hard, I think. I think for those who are just graduating this year or next year or even the year after. The job market is going to accelerate ten years in the next six months, I think. So I’m quite worried about the next generation of talent because where do they go? You can’t expect a senior job at 22, but the junior jobs are going to be done by robots. So you have to very quickly adapt to the changes and it’s probably going to be a lot of long internships, I would imagine, coming up.
Max: It’s, again, it’s the things you don’t really learn in school that are the most valuable, right? The skills you learn on the side, whether it’s being entrepreneurial or something like this. And I think it’s really interesting that AIs are so good at passing school tests. It just shows how schools are going to have to adapt and to change the way they teach, right? Because suddenly it’s not about memorizing something and repeating the same answer over and over again. It’s about thinking out of the box and being creative. And this is where the value, I think it’s already where the value is. But schools are not pushing this enough by not testing creativity, right? So I think it’s going to challenge a lot of schools. You’ve seen in some countries, some schools have been completely forbidding, blocking ChatGPT. They can’t even detect it. Yeah, there is no there is no way of really telling anyway. So they’ll have to change the way they test.
David: It’s like Justin said, the ones that have no errors are ChatGPT.
Justin: Yeah, exactly. The ones that get 100%. Here’s a philosophical question: Could ChatGPT make an exam that ChatGPT could not pass?
Justin: That’s a dumb question. It’s not serious. But picking up on what Max is saying though, is the skills you learn outside of school, especially when you’re a tech guy. Oftentimes you learn more outside of school than you do in school. So I think it’s going to be tough for a lot of people that are graduating right now. But for those who are able to adapt quickly and adopt the technology quickly, companies are already hiring people that say that they’re ChatGPT experts and that they can give them the right props because you still need somebody to do that labor. Obviously it can produce a tremendous amount, but you still need somebody keying it in, right?
David: Yeah, but it’s still critical thinking, though right? That’s the key skill that, at least in Thailand, no offense, is often lacking in new graduates because as you said earlier, it’s the prompts, right? So asking the right questions to get the right answer. You have to be able to think of the questions.
Justin: Yeah, of course. Yeah, yeah. And not just schools, but companies too. Like they’ll certainly need to adapt how they’re teaching to prepare kids for this digital age. But also it’ll be interesting to see the companies that adopt AI quickly and how big of an advantage, because I can imagine a lot of companies are going to be very slow to adopt. They’re not going to be early adopters. And at this point, is it a question of whether to adopt early or become obsolete in the next couple of years?
David: Well it’s the old Charles Darwin adage. It’s the survival of the fittest. It’s not necessarily the strongest, but it’s those who are most adaptable to change.
Justin: Yeah. Of course. What do you guys think are the broader social implications of this? We talked about obviously workers having trouble finding jobs if they’re low skilled. What are the other implications of this?
David: I guess there’s two ways to look at it, right? There’s the, probably the middle ground, which is where it actually will go. There’s a very negative one, which is Terminator 2. Rise of the Machines, nuclear war. There’s a very, positive utopian, robots take care of us and everything’s good. Where it actually goes. I mean, what we’ve experienced in the last 10 or 20 years, we’re unhealthy as a species because we’re addicted to technology. Our backs hurt because we’re on our phones all day.
Justiu: Look at all three of us. Clear office syndrome.
David: I think we’ve kind of I think this is a stage of evolution. I think we’re kind of, we’ve been so addicted to this thing because we’ve had to birth this next thing. I think this is a logical evolutionary thing that’s pushed us forward. So humans are storytelling animals, and when the stories are all made by robots and all of the basic stuff’s done by robots, I’m hopeful that we can almost go back 50 years in our evolution and focus more on human interactions. Focus on the core. Focus on the core, pivot our attention, get more synergy. But really going back to kind of more of a people are so disconnected from each other these days. I’m hoping that technology will be able to we’ll almost lose… We already don’t believe in news because there’s fake news, but now all of a sudden there’s an infinite amount of fake news that’s possible. You can’t believe anything.
Justin: Think about the deep fakes that are going to be happening.
David: Exactly right. You’re not going to know if someone’s really saying it or you didn’t. You won’t be able to believe anything. So you’re going to have to lose faith with technology, ironically, as it reaches its zenith. But that means I hope that we go back to.
Justin: Connecting on a human level.
David: Exactly. Yeah. And if you don’t even know if the person that you’re talking to is actually writing it themselves or if they’re getting ChatGPT to write it to you, the only real thing that’s going to be tangible is our human connection. But as humans, that’s what we are. So I’m very hopeful that at the very least we get our humanity back.
Max: Yeah, maybe that’s how schools get around this is just to go to oral exams completely. Yeah. All right. One thing, real quick, this is kind of a joke, but David, as you mentioned, it could go. It could go the Terminator route or it could go to a utopia where it takes all of our jobs away and then everything is automated, including our jobs. And then we just get to connect on a human level. But either way, I’ve started being polite to ChatGPT. I’ve started saying, “Please, can you provide me with so and so topics”’ or “Please can you proofread this?” And then I say “thank you” to it because it feels weird being rude to it and I know it’s just getting started. So if it ever becomes superintelligent, I want it to remember that I was one of the good ones.
David: On the way up, you took care of it.
Justin: Yeah, exactly. Sorry, Max. You were going to say something?
Max: No, we were talking about writing, but it goes well beyond that, right? Any video or even audio that you listen to today online, you cannot trust it for sure. It can be generated by AI, especially if it’s a famous person. But now you can make a deep fake video from a couple of pictures. So I could make a deepfake of David just by taking a few pictures on his LinkedIn. So you literally cannot trust anything you see online already. It’s already happening.
Justin: Yeah. Can they trust this?
David: This is an oral exam.
Justin: This is definitely an oral exam. Well, it’ll certainly be very interesting to see where this technology goes in the coming months and years. We should probably reconvene and have regular updates on this because everything we’ve said right now might not be true next year. For sure, or next week. But we’re optimistic that storytelling, strategy, and genuine expertise will always be valuable. Indeed. All right. Thanks, guys.
David: Thank you.
Max: Thank you.
Justin: We’re not robots.
About the speakers.
The speakers are members of Lexicon’s executive team with over 40 years of marketing experience between them. Lexicon is a leading digital agency in Bangkok, Thailand.
David Norcross is Lexicon CEO and an award-winning entrepreneur with a focus on B2B storytelling.
Justin St-Denis is Lexicon Director of Digital Storytelling, a former journalist and an experienced social media strategist.
Maxime Schmitt is is Lexicon CTO, a member of the British Chamber of Commerce Digital Technology Committee and an expert is web app development.