All Episodes
Jan 17, 2025S1E400:38:03

4. AI Design + Illustration | Nicola Pringle

Music from Uppbeat: “Make It Happen” by All Good Folks. License code: 6BVFZAHAWVMGPFRE.

Or listen on
Pete BishopHost

Pete Bishop

View bio

Carrie RobinsonGuest

Carrie Robinson

Co-host (Season 1)

Nicola PringleGuest

Nicola Pringle

Art Director, ZGM

Associate Creative Director Nicola Pringle discusses how AI tools like DALL-E and Photoshop's Firefly are integrated into the design and illustration process. She explains that AI serves as a powerful assistant for ideation, concept visualization, and creating client-facing drafts, dramatically speeding up the workflow. While AI has improved significantly, she emphasizes that the final, polished product still requires a skilled human touch to customize and perfect.

How are creative professionals using AI in design and illustration?

Creative professionals use AI as a powerful assistant to explore ideas, visualize concepts, and accelerate the design process. Rather than replacing human creativity, AI serves as a tool to more effectively show an idea that is already in a designer's head.

For designers like Nicola Pringle, AI platforms like DALL-E and Adobe Firefly have become integral to the creative workflow. The process for developing a concept, such as the logo for the 'Artificial Breakdown' podcast, now starts with exploring ideas in AI instead of using a traditional pen and sketchpad. This approach allows for the rapid generation of diverse visual avenues, presenting options that a human might not have considered on their own. The greatest advantage is the ability to quickly create a visual representation of a concept. Historically, designers would use rough sketches and reference images to communicate a concept's style and composition. Now, AI can combine those elements into a single, cohesive image or a series of storyboard frames in minutes. This is a massive time-saver, particularly in a fast-paced agency environment. AI is used to produce drafts that, while not the final product, can blow a client's mind with their realism and proximity to the finished piece.

Which AI tools are most effective for designers?

The most effective tools often depend on the specific task, with DALL-E being preferred for conceptual exploration due to its stylistic range, and Adobe Firefly valued for its seamless integration with Photoshop. Many creatives' first exposure to AI was through efficiency features within the Adobe Creative Suite.

Long before the rise of text-to-image generators, designers were using AI within Adobe Photoshop. Features like automatic image expansion, complex object selection, and content-aware fill were early forms of AI that delivered major efficiency gains on painstaking manual tasks. These tools have steadily improved and introduced many in the creative field to the power of AI. Today, the toolset has expanded. Nicola Pringle primarily uses DALL-E (via ChatGPT) and Adobe Firefly. She finds DALL-E offers the best range of expression and style, making it ideal for the initial concepting phase where exploration is key. Firefly's main advantage is its direct integration into Photoshop, which streamlines the workflow when a generated image needs to be taken further for more detailed, custom work. For the 'Artificial Breakdown' logo, for instance, exploration happened in both tools, but the a final graphic was custom-made by hand in design software.

Does AI replace the need for the 'human touch' in creative work?

No, the human touch remains essential for producing a final, polished creative asset. While AI can quickly generate impressive concepts and iterations, the final product almost always requires human intervention for customization, refinement, and quality control.

Even though AI-generated images are used to develop ideas, it is rare for an agency to use a raw AI image in a final campaign without significant modification. The logo for this podcast serves as a perfect example: it was inspired by AI explorations and early iterations were made by AI, but the final graphic was completely handmade by a designer pushing pixels to get it perfect. The AI gets you to the endpoint faster, but the human touch is what finishes the product. This human element provides a sense of trust and grounding that AI lacks. The imperfections in human art are often what give it value, whereas imperfections in AI art, like a person with six fingers, are seen as errors or 'hallucinations'. AI is not meant to replace the innately human attribute of creating resonant art; it is a tool to assist in that process.

What are the current limitations of AI image generation?

A significant current limitation is the inability to maintain character consistency across a series of images. When generating storyboards or multiple scenes, AI struggles to reproduce the same character without introducing small, unintended variations.

When a designer tries to create a sequence of images featuring the same person, AI often redraws the character from scratch in each new prompt. Even with specific parameters, it is difficult to instruct the AI to take a character from one scene and place them in another without it changing their appearance slightly. A character's hair might change style, their sweater might be different, or other small details will be altered between frames. This is a known challenge that requires a caveat when presenting AI-generated storyboards to clients. Designers must explain that these inconsistencies are a result of the tool and not intentional creative choices. While some platforms like MidJourney are developing features like image referencing to address this, the process can be cumbersome and doesn't always work perfectly. It often requires starting a fresh thread with the AI and re-using the original prompts to get close to the desired character again.

How should people feel about the future of AI?

The general feeling is one of excitement about having a wealth of new resources, but this is coupled with caution about the unknown and the rapid pace of development. For creative professionals, AI is seen as an invaluable assistant that enhances, rather than threatens, their core skills.

The rapid advancement of AI can be both amazing and frightening. On one hand, it's an incredible tool that saves time on iterative tasks, allowing creatives to focus more on what they do best, such as high-level ideation. AI is unlikely to be better than a person at their greatest strength. It is an unlimited, open-ended assistant that speeds up daily tasks in a way that feels as transformative as the arrival of the internet. On the other hand, there is anxiety about what you don't know. The exponential growth of AI since 2022 raises questions about how far it will go and if it will eventually surpass human intellect on a larger scale. The best way to combat this fear is to start using the tools to understand both their capabilities and their limitations. The most powerful tools can get into the wrong hands, leading to issues like deepfakes and misinformation, but for now, the primary effect has been to burst open a world of accessibility and creative potential.
“AI is going to give me a shortcut to showing what's inside my head.”Nicola
“The final graphic is completely handmade, custom made, like by me sitting there, you know, pushing the pixels around to get it just so.”Nicola
“Nothing is precious with AI, which is great. It allows us to iterate more and come up with ideas for longer.”Nicola
“I don't think [AI] is ever gonna be as good at the part of my job that I'm the best at.”Nicola
“You take delight in finding mistakes in AI because it's a bit of a relief to some degree. It's like, that person's got seven fingers. That's so funny. Stupid AI.”Peter Bishop
“If you're worried about it, then just start using it because it's anxiety around what you don't know, right?”Peter Bishop
“It's like an assistant, and it's like an unlimited open-ended assistant in the way that we want to use it.”Nicola

Key Takeaways

  • AI is primarily used in the creative industry to help visualize and show an idea, acting as a shortcut to represent what is in a designer's head.
  • Tools like Adobe Photoshop have been incorporating AI features for a long time, helping with tasks like image expansion and creating clipping paths.
  • For concept work, DALL-E provides a wide range of styles and expressions, while Adobe Firefly streamlines the workflow directly into Photoshop.
  • AI allows creative teams to iterate on ideas more freely and for longer, as nothing is considered precious and concepts can be generated quickly.
  • Despite advancements, AI-generated images for use in campaigns are rarely used without significant human customization.
  • A current limitation of AI is maintaining character consistency across multiple scenes, often resulting in minor changes to appearance.
  • The human touch remains invaluable for the final product, providing a level of trust, quality, and intention that AI cannot yet replicate.
  • Creatives should view AI as an assistant that handles time-consuming tasks, freeing them up to focus on their core strengths like ideation and strategy.

Frequently Asked Questions

How do professional designers use AI in their workflow?

Professional designers use AI primarily as a tool for ideation and visualization. It helps them explore different creative paths quickly and create representations of their ideas to show clients, replacing slower, traditional methods like hand-sketching and creating reference boards. AI acts as a shortcut to create storyboards and concept images much faster than manual methods allow.

What are the best AI tools for creative concepting?

According to Nicola Pringle, DALL-E is excellent for concept work because it offers the most range in styles and expression. Adobe Firefly is also a key tool, as it integrates directly with Photoshop, streamlining the process of moving from AI generation to detailed, hands-on design work.

Has the quality of AI-generated images improved over time?

Yes, the quality of AI-generated images has improved significantly. A year ago, it was common to see mistakes like extra fingers, distorted faces, and other oddities. Now, the images are much cleaner and more accurate, making it increasingly difficult to distinguish between real and AI-generated photos.

Can AI create consistent characters for a story or campaign?

Maintaining character consistency across different scenes is a major challenge with current AI tools. When generating a series of images, an AI may introduce slight variations to a character's appearance, such as changes in their hair, clothing, or other details. This is a known limitation that must be communicated to clients when presenting AI-generated storyboards.

Does AI replace the need for human artists and designers?

No, AI is not seen as a replacement for human designers but rather as a powerful assistant. The final, high-quality creative output still requires the human touch for customization, refinement, and ensuring the work is just right. AI is best at helping with the iterative and explorative parts of the process, but the core creative idea and final execution remain human-led.

How does AI impact the client presentation process?

AI dramatically improves the client presentation process by allowing designers to create much more realistic and compelling concept drafts. Instead of relying on rough sketches and verbal descriptions, agencies can present clients with high-quality visuals that are very close to the final idea. This helps clients better understand the vision and provides a huge time-saving advantage.

What are the ethical concerns about AI copying artists' styles?

A significant ethical concern is that AI can be prompted to mimic the distinct styles of famous artists. While this could be seen as an infringement, the responsibility lies with the human user making the choice to copy another's style. Since people already copy styles without AI, it complicates the issue, but AI makes it possible on a much larger scale.

What were the first AI tools designers started using?

Many designers were first introduced to AI through features built into Adobe Photoshop. Tools for tasks like image expansion, content-aware fill, and creating complex clipping paths for things like hair or trees were early examples of AI improving efficiency long before generative image models became mainstream.

Topics: Generative AI, AI in Design, AI Illustration, Creative Process, Concept Development, Ideation, Storyboarding, Character Consistency, Ethics of AI Art, Future of AI, AI Tools, Marketing

Tools mentioned: Photoshop, Adobe Creative Suite, Firefly, DALL-E, ChatGPT, OpenAI, MidJourney

Carrie (00:49) welcome to Artificial Breakdown. Today we're chatting with Nicola Pringle. She is the Associate Creative Director at ZGM Modern Marketing Partners, and we're very excited to have her on. anybody who doesn't know, she is the one who designed the Artificial Breakdown logo. You're an Artificial Breakdown star. Peter Bishop (01:06) Mmm. haha Nicola (01:11) I'm honored. Thanks, yeah, that was a super fun AI project to get to work on. Absolutely. Peter Bishop (01:12) Yeah, we're pretty proud of that. It looks great, I think. Yeah, I forgot you used AI to design it, didn't you? Or some elements of it anyways. Nicola (01:30) yeah, had to. Definitely kicked off that process with exploring through AI and kind of just trying to visualize ideas by using AI versus the old pen and sketch pad traditional way of doing that. it felt like that's a huge purpose of. what this is about is exploring AI. So it just seemed like the only path really. And it's just so cool. It gives you so many different avenues to consider really quickly things you wouldn't have pictured yourself. Yeah. Carrie (02:03) Yeah Yeah, that's for sure. We've seen some of the images that you created while brainstorming this logo, which we'll probably post a few of those to our socials. but some of them were like, my favorite is two computer faces talking to each other. And I still want that sweater that she's wearing. If somebody could make that, that'd be great. Nicola (02:24) Yeah. Hahaha! Yeah, I think sometimes AI produces exactly what we're thinking of and then sometimes it produces something that you could have never thought of and I think that's just like the brilliance behind it is it's gonna give you things you recognize and it's gonna give you a new way of combining elements and ideas and styles and that's exactly what that was. That concept was like, you know, I had an image in my head of what it should look like and then it did give me the two people with computer screen heads. That's what I was asking it for. And then it decorated it in a way that sort of made it feel familiar. It's just like, it's neat how AI works in a way that actually makes us feel more comfortable with it in certain ways. It shows us kind of a mirror. And that's a nice bridge, I think, between the human. Peter Bishop (03:07) Okay. Nicola (03:23) and tech side of AI. You Peter Bishop (03:25) Yeah, I wanted to dive into that, maybe, Nicola, it might be good just to chat, like back up a little bit as far as you were one of the earlier adopters, I think, here in AI. So it might be good to talk about, what was your first introduction to it and even what you felt at the time and what kind of drew you into it? Nicola (03:45) Well, there's so many things we do in the art department here at ZGM in Photoshop. And I think that's probably the tool that has introduced all of us, like any creative person who's using the Adobe Creative Suite. It's been using AI for a long, long time. And I think basically Photoshop... has been the tool that has grown the most and kind of introduced everybody to sort of the basics of AI. And still to this day, we use it heavily for that. So like, you know, the first things we would have been using it for would be image expansion, you know, a clipping path. So cutting out something out of an image takes forever to do it really nicely. You know, a person with hair, this, you know, kind of... a simple hairdo or a tree with branches and leaves. There's just like many things that are painstaking to do manually. And so I think that's sort of the first introduction to AI is all the improvements and efficiencies that Photoshop provides like that. But then I think Firefly and... DALL-E, know, of ChatGPT and open, like the whole OpenAI suite, we're just starting to just, I mean, we've been using it for a couple of years, these tools, but we're just starting to really see like the full potential. So I personally use DALL-E in ChatGPT and Firefly the most. Firefly is an Adobe. as well. And so that's really nice as it streamlines right into Photoshop when you want to kind of get into more detailed work. And then think DALL-E is the one that gives you the most range of styles and expression and kind of is the best for concept work. So yeah, when it came to the artificial breakdown logo, I was kind of exploring in both areas. within the Adobe Creative Suite and then for those sort of concept images that are really fun and reflect back at us our own lives, that's DALL-E putting that together. Carrie (05:57) Hmm. you find that it's better at having that sort of mirror? Nicola (06:02) I think so, yeah. I mean, there's certain things that it won't do. You know, it's like a reflection of culture and society and so kind of lots of things that we struggle with in our regular world, like representation and, just sort of showing that range of what real people are like. It still sort of struggles with that. You know, it's pulling from the most common things it sees and... you know, it's drawing from images of sort of typical looking humans. And so that's a sort of a specific challenge. But I think in so many ways it shows us who we are. It's like a really neat thing about it. Right? It's just collecting all of our data about us, you know? And then it's like showing it to us. Peter Bishop (06:41) Super weird. Yeah, did you find? Yeah, yeah, it's humbling. Do you find that it's changed a lot over the years? Like as far as what you're getting out of the quality? it gotten? Is it like was it just trash at the beginning? Or is it just kind of the same? Just you're getting better at prompting? Nicola (07:11) yeah, I think it's a bit of both, but for sure it's getting better and better. Like for images of people a year ago, you'd see six fingers or an extra arm in there. Every third person in the crowd would have a smushed face and be kind of strange looking. And it's doing less and less of that now. I think it's harder and harder to tell. what's real and what's AI. And it's just cleaner. it's creating cleaner images. So sort just making less mistakes like that overall. Yeah, I mean, just yesterday when we were looking at the brand assets for artificial breakdown and kind of going through some of the pieces of it, you could see the improvements in the use of that tool. You know, I think it was Firefly creating some images and just how clean the... images of people were and how accurate. And that's something that for sure the first sort of year of everyone using AI for images and to recreate people, that's really, really changed. It's gotten a lot better, which is great and also kind of scary. Carrie (08:18) also kind of scary. Yeah. And something that I find really, I've always found really interesting is people who, maybe didn't want to use AI tools for design, like, well, it's really imperfect, but then the same thing could be said about a lot of human art, which is what gives human art that humaneness is that it is in fact imperfect. So I, like, I want to know what Do you think the differences between an imperfect human created thing that we enjoy and an imperfect AI generated thing that we're like freaks us out. But then also now we have this new era of perfect AI coming out, know, borderline perfect images every once in a while. Like, and those are almost even weirder, but also great. I don't know. I'd love to hear your opinion on that. Nicola (09:14) Yeah, I think it's such an interesting thing. I think we're sort of afraid of AI having the ability to create something so perfect or something that involves an idea that feels so like an innately human attribute that now a computer is spinning back out at us in a similar way and 10 times the speed. But I think. That's part of the journey is like trust and sort of figuring out where we want that and where, maybe steps on some toes or isn't the place to use AI and, propel ourselves forward with that. Cause I think the human touch is still always going to have value and trust. You know, I think that's the biggest piece is like when it comes directly from a human. I mean, I guess that's not really true. I can't trust everybody, but there's something about it that just feels a little more grounded and yeah, real, because it is. And I think, I don't know, I think AI is, I don't know, it's not supposed to be used for that kind of, it's not supposed to replace that part of life, you know? Carrie (10:11) you Hmm. Mm-hmm. Peter Bishop (10:30) It's weird. find it's it's almost like you take delight in finding mistakes in AI because it's a bit of a relief to some degree. It's like, that person's got seven fingers. That's so funny. Stupid AI because the opposite is like, otherwise everything's terrifying. It's like, it's so good at so much of the stuff and what they call them hallucinations or whatnot. Is that, is that right? But it's like, yeah, it's like an eyeball on set of stairs in the background or something and you're just like, wow, that's so so stupid. It's really like a small tiny thing in this work of art that is produced in seconds. But yeah, I was curious a little bit around process, where does AI fit in the grand scheme of, you know, building creative concepts? Is it more upfront in the ideation phase? Is it more on helping you with the finishing side of things or more on quick projects? more on projects that may or may not have a huge budget. I'm just curious, where does it slot in in your day-to-day? Nicola (11:30) Yeah, I think the biggest way that I use AI, and I think it's probably the same for most of our artists in the agency, is to help show the idea. And so you might explore some different ideas or different sort of thought paths through AI to help move that idea process along. But I find the thing I use it for the most is to show a representation of the idea. And so could never draw storyboards for a video idea as fast as AI can create those scenes. And same with an illustrative concept or something that's just very visual. AI is going to give me a shortcut to. showing what's inside my head. And what we would have done before is we would done some sketches, you kind of show, here's what I'm thinking of what you'll see inside of this image or concept, and here's what the words will be, and I sort of sketch it into a thumbnail. And... Then you would combine that with some reference images. It'd be kind of in this style, you know, or this photographer's look and feel would be what we'd be going for. Now it's just like, let's just put that all together. I can show you that in one image or in a series of images for a video or whatever it is in no time. So I think, yeah, that's the biggest advantage and the most key piece of the process. AI for creativity in our agency. We do use it for like actual image use, like to create images we're going to use in a campaign, but I would say it's really rare that create something that we don't then customize quite a bit further ourselves. So, you know, even our logo for this podcast, that is quite AI inspired. And there's iterations of how it came to be its final graphic that were made by AI. But the final graphic is completely handmade, custom made, like by me sitting there, you know, pushing the pixels around to get it just so. So. It gets us to the end point quicker, it's the shortcut, but it's for sure not, you we still need the human touch to get that finished product that we put together here. Carrie (13:49) Yeah. And that's so nice because then you're able to say, okay, I have this idea in my head and you have a client who has, isn't going to decide what to do until they see something, but you don't have to put as much into just this like draft basically, but a draft can blow a client's mind now, because it's so much more realistic and so much closer to what it might actually end up being. but then you actually get your expertise to come in and finish at the job. Nicola (14:18) Absolutely. We're always describing what the final product will be. so AI just helps us to do that so much better. It's not going to be exactly what the thing is. And so we'll have caveats as we describe the idea as you're looking at the image. But man, does it ever give you a leg up on time. And as you most ad agency people know. We have to ride the waves as they come and so we're quite often pressed for time and like it's just awesome. It's such a helpful tool to get us there. Carrie (14:46) You Well, and what I really like about that too, is that something like say you're going to be showing three concepts to a client. That's if you end up building out a full three concepts, the amount of time that that takes and you're only going to actually end up using one, It's, it's so nice in that moment to use that on something that may or may not get created. So Nicola (15:11) Exactly. Nothing is precious with AI, which is great. It allows us to iterate more and come up with ideas for longer, whereas we'd have to cut ourselves off sooner just to be able to put the presentation together before we've had this tool. Peter Bishop (15:17) haha think that's a nice, it's a nice thought to think that it's actually helping us be more creative versus stifling it. Cause I think there's a lot of that sentiment going around where it's like, well, the true creators are becoming less valued because everyone can make stuff now, but there's still an idea behind everything. Right? So I think that's a really nice, thought for sure. I, do you mind if I get super nerdy for sec? There's something that I've been wanting to ask you. So I'm just going to ask you in the middle of the podcast, but Nicola (15:59) Hahaha Peter Bishop (16:01) When it comes to things like character development, I find it challenging because AI kind of redraws everything every time. And I think that's part of it, right? Like it's really hard for me to hang on to, let's say, a look of someone specifically like, I like that, the way that that person's looking. I want to put that person in different scene and, and make like a series with that person. And I find it really hard to hold on to that. what it generated because it keeps regenerating everything completely different almost every time even though you give it the same parameters. you have? Is this a thing? My like, way out to lunch? Or is that a challenge right now? Carrie (16:39) You Nicola (16:39) Hahaha! No, I think that is a challenge. think it would be so great if, let's say, you're putting together a storyboard for a video. And you could say, create your first scene. And you've got that looking good how you want it. like, OK, now take this guy in the red shirt from this scene and have him walking down the highway in a rural area or whatever. And it will. bring some of the previous scene with it, or it'll put that rural landscape, if you say landscape, it probably put it in a painting in the same scene you were just in or something, you know what I mean? And it's like, no, do this, give it another prompt. Yeah, you give it another prompt to sort of help get to what you want. And it sometimes just hits a dead end. And it'll just kind of keep spitting out the same image saying, hey, here's your revisions. Here's the edit you asked for. And it's like, So I'll often sort of fresh thread. Peter Bishop (17:18) So literal. Carrie (17:19) you Nicola (17:40) with the AI program and like you said give it those prompts that gave you that character in the first place. I don't know. I haven't found a way to take a reference image of a human, like of a person, representation of a person forward through a whole bunch of different scenes. And that is something that is a caveat when sharing storyboards with clients. It's like, you'll notice that we've made these with AI. And our main character here has a slightly different wave in her hair here. know, in this other scene, it's a little bit different. We're sweaters. Yeah, exactly. Yeah. And now our nails are painted. It's like, little things. It's like, we wouldn't change these things. These are just AI. Peter Bishop (18:11) Now she has seven fingers. Yeah Yeah, I've had luck with Mid Journey with that a little bit because they have this image reference, which is actually real pain to use right now because it's not an easy thing to find. You have to do this slash image ref or whatever, and then you have to have a URL to the image. So somehow you got to stick your image up on a site somewhere. it did seem to kind of hold together pretty good, which was kind of cool. Nicola (18:21) know, glitches. Peter Bishop (18:44) those are the things that I think like the, and I've seen it in DALL-E, where there is kind of image references that you can pop in for style and some sort of like visual effects and whatnot, but it does seem pretty loose. Nicola (18:59) Yeah, yeah, you can throw in a reference image or previous rendering to say, you know, match the lighting in this one or show as much of the scene as that one did. It doesn't always work. Yeah, this is true. Peter Bishop (19:14) so you're an illustrator too and I really think, I'm curious what you think about maybe it's more the ethics side of AI when you look at like okay if I'm putting my own illustrations out into the world and someone's grabbing them and potentially making artwork using my style like does that how does that feel does that feel like an infringement does it feel flattering like where where do you where do sit there Nicola (19:39) I think it's both. It's definitely an infringement and I think there's people who have pretty distinct styles and are very famous for them and then they're being copied. They're being copied in so many ways and that could be through DALL-E, a filter in DALL-E that just like that's that person's style. You can't type in a person's specific style but you can describe it and get pretty close because anything popular is gonna mirror back towards you. Peter Bishop (19:41) Ha ha ha ha ha! Nicola (20:10) to create something that already exists. yeah, I don't know. I think that's a tough one because people are doing that without AI already. And so it's kind of hard to say where the fault lies. It's the human being that's choosing that. So I don't know. I don't know how to decide who to blame there. That's like a human has to make the choice to steal that from you, you know? Carrie (20:31) You Peter Bishop (20:35) Yeah. Carrie (20:36) Yeah, cause it is one thing if you see a card on a shelf that you're like, clearly that was somebody else's style that, know, whatever card company stole, but now you have hundreds and thousands and potentially, you know, all these people that can do it. And they're all creating a really similar notebook to somebody else's notebook that they're now selling on Amazon or whatever the case may be. And it's just, it's the sheer amount. that can be copied. Like you said, people are always going to steal people's ideas and work and styles, but yeah, it's a lot this time around. Peter Bishop (21:15) Hehehehehe Nicola (21:15) Yeah, mm-hmm. And AI makes it tricky because, you know, it's not going to be exactly that person's style. There's going to be some variation. And so there are these laws about what percentage of likeness is legal when using other work as inspiration. So I don't know. It's a tricky one. Yeah. Peter Bishop (21:36) Yeah. What are you looking forward to? Are there things that you're hoping that will come out soon or things you've heard in the grapevine as far as new features or tools that you're excited about? Or where do you think this is all heading on your end? And are you excited about them? are you like, my god, am I going to have a job in five years? How are you feeling about the future here? Nicola (22:03) Yeah, I think we just have a wealth of resources. Like, it's just exciting. I know there's so many tools that I haven't even tried that I'd love to try. I haven't really dipped into the animation world. in that way, but even just within our apps that we already use, there's so many tools that I haven't used. But I think my general feeling is I love it. It's just great. think it's like AI, I don't think is ever gonna be as good at the part of my job that I'm the best at. I think that that's the thing about AI is whatever you're the best at in your role. it's likely that AI isn't going to be better. But AI can help you get to that point of showing. If it's, say, ideation, it can help get to that point of showing the ideas and figuring out how to put the visuals together. And I think that's what's so brilliant about it. It's like an assistant, and it's like an unlimited open-ended assistant in the way that we want to use it. Carrie (23:11) I feel like we've heard that a lot and we'll continue to hear that a lot. And I think that's of the directions that AI is going in the workplaces. People using it as an assistant. It's certainly not a replacement, but it's definitely there to help you along the faster. Nicola (23:29) Yeah, and just do that exploring part, that iteration part that you really want the time for. It's just such a time saver in that way. Yeah. Carrie (23:41) Mm-hmm. Peter Bishop (23:41) Yeah, it's funny when you mentioned how Photoshop was kind of the gateway drug. And some of these things, I think it was like content-aware fill, was kind of the beginning of this, where there was like, yeah, you can erase stuff and it will figure out what's behind it. This stuff's been dribbling in slowly, and everyone thinks it's great, until a critical amount comes in, and then suddenly it's scary. we've been kind of working with this all along. Some people say autocorrect was the beginning of AI. whatever it is, but it's just, feels like if, if this was spoon fed out a little bit slower, it'd be like that frog in the boiling water. No one really noticed too long, but we just suddenly got a whole glut of it all at once. Like suddenly chat GPT opened up the everything for everybody instantly. And it was, it's like this massive tool that just became commercially ready so quickly. think that scared people all at once, but if that maybe come out slower, I wonder if it would have been. less of a big deal and people adapted like at their own pace. Nicola (24:43) Yeah, and I wonder who is really afraid of it, who doesn't trust it, if it is people who use it or if it's people who don't use it and they don't really know, you know, the extent of what it offers and its limitations too. Peter Bishop (24:59) Yeah, I think we've been saying this all along. If you're worried about it, then just start using it because it's anxiety around what you don't know, right? Nicola (25:08) Yeah, exactly. Which is, think, why the future of it is scary, because we don't know. And it does seem to be moving so fast. It's like, at some point, is it going to move beyond us? I don't know. I think that's the, that unknown is like the scary part. But right now, it feels like we've just like burst into this awesome. Peter Bishop (25:16) Yeah. Nicola (25:29) world of accessibility to speed up so many of the things we do on the daily. So it's just like, it's like when the internet arrived, just like, holy crap, this is amazing. Like I can see everything. Like it's just, you know, it kind of feels like that. It's like, this is gonna be just huge. Yeah. Carrie (25:47) Yeah, especially when you think like chat GPT was launched in 2022. I'm pretty sure. Nicola (25:55) No way. Really? It's only been that long. Carrie (25:58) like it hasn't been that like, but the exponential growth that it's had, it was, it's like, and technology and Peter and I've talked about this before, but technology grows at more and more of it at like exponential rate. It's like, what that's been two years. What's going to happen in two more years. Like, and then two years after that, two years after that, and it's, it's pretty. Like you just can't, I don't think our tiny little, you know, P human brains can comprehend it. It's. Nicola (26:15) Mm-hmm. Carrie (26:28) It's not within our comprehension. We're not meant to be processing this much information at a time, but yet we're given these tools and we're like, yeah, let's, let's fuck shit up. Nicola (26:41) And the whole goal is just like, let's keep making them better. Let's keep making them smarter. Let's keep making them do things faster. And it's like, what's the goal? How far do we want this to go? Yeah, this feels good. It certainly feels like there's potential for some of those movies about AI to eventually come true, the scary ones. Carrie (26:50) Yeah, what's the actual goal? Like if it's just good, can we just stop when it's good? Do we have to keep going? Peter Bishop (26:58) Yeah. Carrie (27:06) Yeah, well, I wanted to go see there's a horror movie that came out recently about AI. I haven't seen it yet, but Peter and I were talking about going to see the wild robot instead. I'd rather watch that. Peter Bishop (27:06) Yeah. Yeah. Nicola (27:18) Yeah. Yes. Yeah. Peter Bishop (27:20) Yeah, I really yeah, does it does make terminator you look at terminator in a whole new light I'm like, I hope that's not a documentary like Nicola (27:30) and that network takes over. Carrie (27:30) It's funny how long ago in like pop culture that started like it's almost like the second machinery existed. It was like, this could come take over. Like I'm sure there's ones about like steam engines coming and taking over. Like there's things, what's that one? The friendly, the brave little toaster. Peter Bishop (27:51) What? Carrie (27:52) The brave little toaster. That was a kid's movie when I was growing up and it was all these things that were very like humanoid. I don't think they were taking over to be fair. They were just being cute, but... But I feel like humans have always been kind of afraid of that. Nicola (28:02) Hahaha Peter Bishop (28:03) That, So many predictions came true, but the one thing that they haven't got is flying cars and jetpacks. They've been calling those out since day one. Nicola (28:06) It's like if you're a brunette, it's bad. Yeah. Peter Bishop (28:19) I'd trade in my iPad for a jetpack. Nicola (28:19) Yeah. Carrie (28:21) You Nicola (28:22) Yeah, we want the back to the future scene to come true. Come on. Where's my hoverboard? Right? Peter Bishop (28:26) Yeah, hoverboards, yeah. Carrie (28:30) Okay, a hoverboard would be sick. Nicola (28:32) I just saw there was a, and I knew it, this is, you'll cut this out. knew this isn't really what it's like or what it's about, but the Nobel prizes were being handed out recently. And did you hear what the godfather of AI had to say? It was like a warning. It was like warning the world that it's going to develop fast and. basically promising us that it will go beyond human intellect and ability in a lot of areas and that we should make sure we're paying attention and keeping up. Peter Bishop (29:10) Yeah, that's terrifying. Yeah. I think, cause everything we're talking about right now is kind of like the generative AI stuff and it's still pretty much like all controlled by humans. But the second they get into like bots, teaching bots and exponentially getting out of control, then that's where it's like, Holy fuck. Like, I don't know what the answer is there. It's just like, hang onto your socks. and hope you still have any job, even if it's going to be like mining for silicone. Nicola (29:11) Sorry. Right. And they just. Carrie (29:40) That's pretty much it. Nicola (29:41) Right, right? So many things will be machine Peter Bishop (29:42) You Nicola (29:44) driven and machine learning will be like exponentially quicker than human learning of these things. then, yeah, I mean, that's happening already, but it has to happen on a much larger scale. It feels like it should. Carrie (29:58) Yeah. And I mean, he's not the only person to be giving out warnings recently either. There's a couple people who have been saying, like people who work in the creation of AI are, you know, they've got bunker houses that they're ready to go to with their families. And it's like, why do you have that? Nicola (30:15) It is this thing, like, it's a powerful tool. And so it depends on whose hands it gets into. I think that's for sure of kind of a big piece of it is, you know, like the deep fakes that are out on the internet and, you know, how politics can use AI to influence and that part is like, yeah, it's like powerful tool can get into the wrong hands. It's like a Marvel movie. You know? Carrie (30:25) Mm-hmm. Mm-hmm. Peter Bishop (30:40) Ha ha ha ha. Carrie (30:42) Yeah, that's life, baby. You gotta live on the edge. Peter Bishop (30:47) Yeah, and I also think too, it's great link bait right now too, to say that the earth is ending. So I imagine some of these are maybe, maybe that's driving some of them. But if you're actually going through the motions of building out a bunker, then that feels more than that. that's, yeah, that's somewhat terrifying. Okay. Well, on that note. Carrie (31:07) You Nicola (31:11) notes. Yeah. Peter Bishop (31:12) You Carrie (31:13) But that's why you need to listen to our podcast to stay on top of it all. Nicola (31:16) Yeah, keep up, you know, be ready. Peter Bishop (31:18) Yeah, yes, totally. Bunker diagrams are in the show notes. Nikola, thanks for coming on. That was super cool of you. It's great to hear from you, because again, you've been kind one of the designers who started this in kind of day one, or at least certainly feels that way. At ZGM So yeah, I really appreciate your thoughts on this. Carrie (31:25) Hahaha! Nicola (31:41) Thanks so much for having me. It's a really interesting subject, so always happy to chat about it. Carrie (31:48) Great. We'll have you on next season again to see what level we're at of bunkers and or Photoshop. Nicola (31:50) Okay, cool. Peter Bishop (31:52) Yeah Nicola (31:55) Maybe I could get a mic that works and better lighting. Peter Bishop (31:59) is the dream. We'll have self-aware robots, but we'll still be working on microphones. Carrie (32:04) You Nicola (32:05) know, muting and unmuting. It's hard. Okay. Thanks, team. See ya. Peter Bishop (32:12) All right, thanks again. Carrie (32:13) Thanks, Nicola. Peter Bishop (32:16) Bye. Nicola (32:17) Okay, bye. Carrie (32:18) Nicola, you have to leave. Peter Bishop (32:20) Hahaha Carrie (32:31) I'd like to talk about whose guest that was. I feel like she's my guest, but she feels more aligned. Yeah, I am. Peter Bishop (32:33) I didn't know we were keeping tabs. Hahaha Fair Carrie (32:45) She could be a 50-50, I guess. Peter Bishop (32:47) Mm, that's fine. Fine. I think I have some gold coming up. But yeah, she was great. Like, so I think there's so much around the visual side that's creeping up so fast. And it's exciting to hear how artists and graphic artists are reacting to this stuff. So it was great to get her perspective. Carrie (32:50) That's fine. And I really liked that idea of AI is not the one coming up with the idea. AI is not the one giving you a finished product, but it's in the mix in between those two things. And it helps us along the way, you know, Nicholas brilliant. And she comes up with really great ideas and she then uses AI to help her through the process to eventually create something amazing. yeah, I just love that, that genius, that human genius is still there. Peter Bishop (33:36) I remember thinking this a long time ago. I always thought it was almost elitist to think that, okay, only people who can draw or write are creative, right? Like everyone's creative and a lot of people just have different ways of expressing it. But I think of other people in my life who are, just don't consider themselves creative because they can't put an illustration together, but they are. Carrie (33:47) Mmm. Peter Bishop (34:02) They, everyone's has created ways of thinking and, and I feel like in some ways AI takes a bit of that elitism out of the equation. If you know what I mean, it's similar to programmers. again, nothing but respect for programmers. But if I have a great idea for a product, well, now I can build it in some ways, right? Or I can draw it or I can write it. And if it's still a good idea, it's just, I didn't have the skillsets to do the. production on it or the finishing side. So it's just an interesting way to look at creativity, I think. Carrie (34:37) Yeah, which is a funny, you know, intersect back to ZGM with this idea that like we don't call our creative quote unquote department, we call it storytelling, which I like because we're not the creative department. Every department is creative in their own way. Strategy is creative. You know, all these accounts, people are creative in their own ways. And I totally agree. It's like, It allows you to take your creativity and just show it in different ways that you didn't have before. Which is why I like DALL-E as a writer. I can go in and ask it to make something that I wouldn't have been able to create. I would have. Peter Bishop (35:07) Yeah. That's why I like the text ones because they just writes for me because I have the writing skills of a maybe four year old. Carrie (35:25) Great, four year old. There it is, in action. Peter Bishop (35:31) I thankfully I've written my name long enough now that I don't misspell it, but I misspell pretty much everything I write. Have you heard of this? I saw this the other day where you can take any subject as long as it's written out and AI will generate a podcast for you. So you can listen to it like a podcast. So it's grade 12 biology. will. Carrie (35:36) Hahaha Whaaaa- Peter Bishop (35:58) convert it into a play-by-play audio podcast for you so you can digest it easier. Isn't that nuts? With two hosts. Carrie (36:05) Why are we even recording this? Why aren't we just doing that? Peter Bishop (36:09) right hey maybe this is that maybe we're Carrie (36:14) Dun dun dun. The conspiracy theory continues. Are Carrie and Peter bots? Peter Bishop (36:22) Yeah, I would like to think AI is better than this, because this is, I don't know. But yeah, isn't that nuts? Carrie (36:30) wow, that's pretty cool though. Cause again, that's something to do with like, you know, everybody has these different types of learning. Like there's visual learners and audio learners and I don't know the proper terminology for it all, but you know, imagine you're given something and you're struggling to learn it, get AI to turn it into something else. How accessible is that? Peter Bishop (36:37) Right. Totally totally like Yeah convert like even just putting it into an audio file is just great But I love the thought of just like putting it into a conversation Maybe you latch on to a conversation better than you would just a basically a speech Which is what most textbooks would be If you transcribe them into audio But yeah, I think it's really cool sort of highly threatening really Carrie (36:55) Mm-hmm. Yeah. That's how people describe me. That's so funny. Peter Bishop (37:17) Ha ha ha! Okay, I think we're Yeah. Okay, I'll talk to you later. Carrie (37:23) Well, on that note, great guest. You're welcome. sweet. Yep. Bye.