We sat down for a conversation with Colin Dunn, cofounder and CEO of Visual Electric. He'll share his journey from product designer to pioneering the first AI image generator made specifically for designers. Beyond his inspiring story, Colin will offer a live demo of Visual Electric's unique capabilities, showing why it's an indispensable tool for modern designers.
View additional design talks and workshops on our Events page.
Prefer reading vs. watching? Below is an automated transcript of the event cleaned up with AI.
Emilyann:
So, a little bit about Designlab before we get started. I just wanted to share a bit about who we are. Obviously, we're hosting this talk in conjunction with Colin and Visual Electric. We’ll talk more about Colin and Visual Electric as we go through the presentation, but for those of you new to Designlab, I just wanted to share a few words about us.
We are an online educational platform focused exclusively on digital design. Our education model combines project-based learning with expert mentorship, and we've been operating in this online space for over a decade. More than 20,000 students have taken our classes. We have programs for those transitioning into design careers, designers who have been working in UX/UI for some time and want to upskill, and programming for design teams through team trainings. We even cater to people who are totally new to design and may not be looking for a career change but just want to learn new skills.
For today’s attendees, we’re offering a $100 discount code that can be applied to any of our courses, whether you're at a beginner or advanced level. Just enter the code on the screen at checkout, and it will be automatically applied to the course of your choice. I’ll also drop a link in the chat with the coupon attached, just in case you want to check it out directly.
Without further ado, let me introduce Colin Dunn.
Colin Dunn is the co-founder and CEO of Visual Electric, the first image generator built specifically for designers. Before starting Visual Electric, Colin led design at Universe and Playspace, and prior to that, he was a product designer at Dropbox and Facebook, which many of you might be familiar with.
We’re going to start with a Q&A with Colin for about 15–20 minutes, followed by an awesome demo of some of the things you can do with Visual Electric. Then we’ll wrap up with a final Q&A at the end.
Emilyann:
So, to start, Colin, can you elaborate on what you mean when you say Visual Electric is an image generator made specifically for designers? In what ways is this tool created with designers in mind?
Colin Dunn:
Yeah, it’s a great question. First, thanks for having me! I’m excited to be here and show you Visual Electric and answer any questions. As a designer myself, it’s fun to build a product that I would want to use. When image generation technology first came out, my co-founders and I were excited about it, but as we played around with the available tools, we noticed that they weren’t very friendly to the designer workflow. They felt more like tech demos, built in environments like Discord bots, rather than thoughtfully designed user interfaces.
We wanted to build something that aligned with how designers think and work, so the first thing we did was put the image generation technology into an infinite canvas. Many existing products were using chat-based interfaces, which work well for large language models, but for visual work, it felt unnatural. Designers need to work in a nonlinear way, bringing in reference images and experimenting with visual elements side by side. So, we built workflows like collaging, where you can layer images on top of each other and generate from that, and retouching tools designed to fit the design process. We’re trying to create a truly visual workflow for working with this technology.
Emilyann:
That’s really impressive! You saw a need and expanded it in a way that fits the design space. In an interview I saw on YouTube, you highlighted the importance of solving the right problem when starting a business. How did you identify the problem Visual Electric is trying to solve, and why do you feel it’s the right problem?
Colin:
That’s a great question. When you’re a founder, especially early on, you’re constantly asking yourself, “Are we solving the right problem?” I think of it as being like wandering through a forest in the dark—you’re trying to find your way and the right path.
Initially, we were excited about the technology and knew there was potential to build something new in the space, but the exact form that would take wasn’t clear. We took a first principles approach, asking ourselves which types of creative work would be disrupted by this technology. We considered everything from video and music to 3D and game design. It wasn’t until later that we honed in on graphic design and imagery.
We focused on this area because, first, it’s a large market with several established companies, and second, image generation technology was improving rapidly. From a technical and business perspective, it seemed like a great opportunity. The third reason was that it’s something we’re passionate about. Starting a company is hard and takes time, so having that enthusiasm for what you’re working on is crucial.
Emilyann:
I’m sure it takes a lot of discipline, passion, and awareness to create a product and company. Once you identified the problem or opportunity, how did you make the jump from working for someone else to being an entrepreneur? What was that journey like for you?
Colin:
I’ve wanted to start a company for a long time and had many bad startup ideas, which I’m happy to share. I’d been thinking about starting a company for years, exploring different ideas at different levels of development. One person who was really helpful in making the leap was Greg Rosen, a pre-seed investor who specializes in working with founders at the earliest stages—often even before they have a solid idea or co-founders.
Greg and I built a relationship over 10 years, and he gave me feedback on various ideas. When I brought him the idea for Visual Electric, he was really excited about it from a business perspective. He knew I was uniquely positioned to build a creative tool like this because of my background. He ended up writing the first check for Visual Electric, which allowed me to quit my job and work on the company full-time.
Emilyann:
That’s fantastic! It seems like it’s gone really well for you, which is amazing.
Colin:
I’m glad it seems that way from the outside! As a founder, it feels like an uphill battle every day, but I’m glad it looks like it’s going well.
Emilyann:
In that same interview, you mentioned how Visual Electric’s design process has involved rapid iteration and constant user feedback. Could you share an example where user insights significantly altered the direction of a product or feature?
Colin:
Absolutely. We’ve taken a very iterative approach to product development. One thing we’ve done is use a service called User Interviews to get prototypes in front of real customers for qualitative research. A major hurdle with this technology is that while it’s clearly powerful, people don’t want to become prompt engineers. They don’t want to spend time learning the ins and outs of the model or the perfect phrases to use.
Our focus has been on creating an interface that abstracts that complexity away and helps users bring their vision to life without needing to be experts in machine learning. One example is our Creativity Slider, which shows up in different places in the product and allows users to control how much variance they want in the images they generate. While users may not understand the exact mechanics behind the slider, our testing shows that it works intuitively, providing them with control in a way that feels natural.
Emilyann:
That’s really cool! As someone who’s used a lot of AI tools in the market, I can definitely relate to the complexity of some of them. It’s great that you’re working on making this more intuitive for users.
So, as a founder with a deep design background, what were some of the unique challenges you personally faced in transitioning from being a product designer at big-name companies to leading a more design-focused startup? What are the big lessons from your previous work experience that you're applying at Visual Electric?
Colin:
Yeah, when you're a product designer at a large company, your job is primarily to represent design and user experience. You’re advocating for the customer, but you know there are other teams—like engineering and business—advocating for their own goals. When you become a founder, though, you’re responsible for balancing all of those things yourself. You’re no longer just focused on user experience—you’re also thinking about the business model, the cost of developing certain features, and how you’re investing your resources.
As a designer, I’ve had to recalibrate what I think is most important to work on. I’m only wearing my designer hat about 15% of the time now, while the other 85% is spent on things like customer support, user testing, copywriting, marketing, and social media. There are so many other roles you have to play as a founder.
If you’re the kind of person who wants to spend all your time doing design, being a founder may not be the right role unless you have a co-founder to handle those other responsibilities. But for me, building a company is still a design problem—it’s just a much broader design challenge that includes everything from the user interface to the employee experience, to the overall company culture and how users interact with the business at every touchpoint.
Emilyann:
Yeah, I imagine it requires a very holistic outlook and the ability to wear many hats throughout the day. Thank you so much for sharing your insights! I’d love for us to move into the demo now so we can see more of Visual Electric in action. I’ll hand it over to you.
Colin Dunn's Demo of Visual Electric:
Colin:
Perfect! Let me share my screen, and I’ll walk you through Visual Electric.
When you first log in, you’ll see an Inspiration Gallery. These are images generated by members of our community. You can click on an image to see the prompt that was used to create it, as well as additional information like the aspect ratio and other features. From here, you can add the image to a new canvas, copy a link, or download it.
If you click on the Images tab, you’ll see a timeline of all the images you’ve created. Up in the top corner, there’s a workspace switcher, where you can create or join shared workspaces. By default, you have a personal workspace where everything is private to you, but shared workspaces are great if you’re collaborating with a team.
I’m going to create a new canvas now. When you do this, you’re dropped into the editor, and if you’ve used tools like Figma or Sketch, it should feel pretty familiar. On the left, we have the Inspiration Gallery, where you can add images to your canvas, and you can see the prompt that was used to generate each image.
At the top of the screen, there’s the Composer, which is where you generate new images from prompts. I’ll type in “a simple window reflecting a peaceful garden” and start generating. As I type, you’ll notice that we give you suggestions. These suggestions are powered by a model we trained to understand what makes a good prompt. A basic prompt needs a subject and a location, and once you have those, we offer suggestions for additional creative elements you might want to include.
We also have a Color Picker if you want to define a color palette for your image. And then there’s the Exclude section, which allows you to exclude certain elements from your prompt. For example, if you don’t want the model to generate anything with red, you can add “red” here, which will steer the model away from using it.
We also have a Style Library with over 55 hand-created styles. Each style represents a unique way to interpret your prompt, ranging from photography to fine art, illustration, and digital art. I’ll choose a hand-drawn style for this example, set the aspect ratio to square, and keep the quality at the highest level.
Once you start generating images, they’re added to the canvas. You can select an image, and the right-hand sidebar will show more options for working with it. I’ll generate a few more images in different styles to show you the range. For example, I’ll try flat design and Portra 400, a popular analog film style. As you can see, with the same prompt, you can create very different images just by selecting different styles.
One of the things our model does really well is realistic photography. This is great for replacing stock photography, which often feels overused. You can create images that are on-brand and unique to your project. I’ll show you how to do that in a moment.
Another cool feature we have is the ability to work with found imagery. If you have an image you like, you can simply drag and drop it onto the canvas. From there, you can click Generate Prompt, which will analyze the image and generate a description that you can then modify or use to create new images. This is especially helpful if you find inspiration from existing artwork or photography and want to create something new in a similar style.
Emilyann:
Wow, that’s super impressive! I’m really excited about how user-friendly it all seems. I love that you’ve made it intuitive for people to generate images without having to go too deep into the technical side of AI.
Colin:
Thank you! I’m glad it feels that way because we’ve spent a lot of time iterating to make it simple and intuitive.
Emilyann:
Now that we’ve seen the demo, can you give us some insight into the future of Visual Electric? Where do you see the product evolving?
Colin:
Definitely! We’re really focused on serving designers and brand designers specifically. Right now, Visual Electric is great for replacing stock photography and creating unique illustration sets, but one of the things we’re working on is the ability to generate images that include specific products, logos, or characters from a brand. This is a huge use case, especially for brands that want consistency across their imagery.
We’re also working on making the product even more accessible. That means allowing users to generate images outside of the canvas environment, which can open up more possibilities—especially on mobile. We know people want to use this tool on mobile, so we’re working on that as well as developing a mobile app.
Emilyann:
That’s really exciting! I also have to say I’m impressed with the text generation in your tool. That’s been a challenge for a lot of AI apps, but yours seems to do it quite successfully.
Colin:
Yeah, our model is great at generating text, although it can take a few tries to get everything perfect. One thing we’re working on is integrating all of our tools, like retouching and inpainting, into our newer models to make these processes even smoother.
Emilyann:
That’s super exciting! It sounds like there’s a lot to look forward to with Visual Electric. Is there anything else you’d like to share with us before we dive into the Q&A?
Colin:
Yes, for those of you attending, I wanted to offer a discount code. You can use the code Designlab for $20 off, which gives you a free month of our standard plan. I’ve put the code in the chat. It’s good for first-time customers, and you can generate 20 images for free every month. With this discount, you’ll get an extra 500 fast generations for the first month.
Emilyann:
Amazing! Thank you so much, Colin. I’m sure many people here will take advantage of that. And for those interested in Designlab courses, don’t forget we’re offering $100 off for participants in this talk as well.
We’ve got quite a few questions coming in, so let’s jump into the Q&A.
Emilyann:
Here’s a question from Katriana: “If you choose the fastest speed for generating images, can you go back later and make it high-quality?”
Colin:
That’s a great question! I wish you could, but unfortunately, that’s not how the technology works right now. The faster speed uses a smaller model with fewer parameters, which is what makes it faster. We are working on an upscaler, which will allow you to take a lower-quality image and generate a higher-resolution version of it. We have an upscaler in the product now, but it’s based on an older model, and we’re updating it to work better. But for now, you would need to regenerate the whole image using the higher quality setting.
Emilyann:
Great, thanks for that answer! Paul says: “I’ve been using Visual Electric for about a year, and I love it. One of my favorite features is the ability to develop your own style. But I find it difficult to get it just right. Any tips?”
Colin:
Thanks for being a Visual Electric user, Paul! I’m glad to hear you’ve been enjoying it. Developing custom styles can take some trial and error. My first tip would be to use the Create Style button on the bottom right when you have an image you like. This lets you build a style based on that image’s prompt.
Creating styles, especially ones as unique as the 55 styles we have in the product, takes a lot of iteration. My advice is to try different words or phrases to describe what you’re aiming for and experiment with the order of words in your prompts. The model reads the words in order, so the most important keywords should come first. And don’t hesitate to reach out if you need help—I’m happy to assist with creating custom styles!
Emilyann:
That’s really helpful! Andrew wants to know: “Do the images generated have any metadata or indicator that they were made with AI?”
Colin:
Yes, that’s a great question. Each image generated by Visual Electric has an invisible watermark embedded in it, which isn’t detectable to the human eye but can be used to verify that the image was AI-generated. This is becoming increasingly important as AI-generated images become harder to distinguish from real ones.
Emilyann:
Very cool! Another question: “As a designer using images from Visual Electric in my projects, what do I need to do to ensure I’m following guidelines for image use? Are there specific steps for crediting or securing permissions, both now and in the future?”
Colin:
Great question. If you're on the Standard Plan, we grant you a personal use license for any images you create. If you want to use an image for commercial purposes, you’ll need to upgrade to the Pro or Ultra tier, which includes a royalty-free commercial license. There’s no attribution required, and this license applies to all the images you’ve created—even the ones you generated on the Standard Plan—once you upgrade. The commercial license is also in perpetuity, meaning you don’t have to maintain a subscription to keep using the images commercially.
Emilyann:
That’s awesome to know! Moving to a broader topic—what are the things you wish you knew before you became a founder? Any advice for future founders?
Colin:
Oh, there’s a lot I wish I knew! One of the biggest learning curves for me was fundraising. It’s both an art and a science, and I could probably spend an hour just talking about what I’ve learned. But the biggest lesson is to be in control of your own fundraising process. When I was starting out, I got caught up in all the inbound interest from investors and let them dictate the pace and direction of my fundraising. In hindsight, I should have paused and waited until I was ready to raise on my own terms. I recommend being clear about how much you want to raise, talking to as many investors as possible, and running your own process. That way, you stay in control of the outcome.
Emilyann:
That’s such valuable advice, thank you! Kinda has a question: “Are there any situations where it could be considered unethical to use a tool like Visual Electric for AI generation?”
Colin:
That’s a great question and one that’s on a lot of people’s minds. There isn’t a consensus yet on the ethics of AI, and different people feel differently about it. As a founder of an AI company, I’m obviously very supportive of the technology. I believe that making creativity more accessible is a net positive for the world. However, I also understand that for people who make their living doing work that AI might replace, this is a scary development.
I think back to when the camera was invented—painters were concerned it would put them out of work because, at the time, portraiture was a huge part of their business. But what happened was that photography allowed painting to evolve, giving rise to new art movements like cubism, abstract expressionism, and others. In the same way, I see AI as a tool that can open up new creative possibilities. I don’t believe it’s a zero-sum game, but I do recognize the anxiety it causes for some people.
Emilyann:
That’s a really thoughtful answer, Colin. We probably have time for one last question: M. Kind wants to know, “How much did you know about the technical side of AI before starting Visual Electric?”
Colin:
Honestly, zero. I was completely ignorant about how machine learning worked before I started this company. I’ve had to learn on the go, and I’ve been fortunate to work with smart people who understand this technology. My co-founder, Adam, is a machine learning engineer and the domain expert on our team. We work closely together to design the product experience based on the technology’s capabilities, and I’m constantly learning from him. I wouldn’t call myself an AI expert by any means, but I now have a deep enough understanding to contribute to the product design and development process.
Emilyann:
That’s really encouraging for anyone considering venturing into AI without a technical background. Thank you so much, Colin, for joining us today and sharing all your insights and experiences! It’s been a fascinating discussion, and I’m personally really impressed with Visual Electric’s capabilities.
Is there anything else you’d like to add before we wrap up?
Colin:
Thanks again for having me, Emilyann! I’d just like to remind everyone to follow us on Instagram and Twitter. Feel free to reach out—we’re very responsive, and if you sign up for a paid account, the little chat box in the corner of the app goes straight to my inbox, so I’m happy to help with any questions.
Emilyann:
Amazing! And don’t forget, everyone, the discount code Designlab gives you a free month of Visual Electric’s standard plan, so make sure to grab that while it’s available. Thank you again, Colin, and thank you to everyone who joined us today! We look forward to seeing you in Visual Electric and at future Designlab events. Have a great day!
