Interview with Amy Balliett, Visual Strategist
Q&A with Amy Balliett, who is presenting at The Design + AI Summit 2025
Amy Balliett is an author, speaker, and visual strategist. She’s also speaking at The Design + AI Summit, being held online November 13-14, with a sessions on AI Agents in Action.
We thought it would be fun to get to know her better with some Q&A.
Biggest mistake designers often make when using AI?
Assuming AI will replace designers or is an enemy to their profession and therefore avoiding it. Highly skilled designers will never be replaced by AI, especially if they learn how to use it to their advantage. AI can be a powerful design assistant, adding efficiency to the design process while making it far easier for designers and stakeholders to align on a vision.
Sure, AI might produce some eye popping visual content, but it won’t produce perfect final drafts with any ease or consistency. To take AI content to the next level, a designer skilled in the Adobe Creative Suite is necessary. Making the most out of text to image AI requires a great creative vision and the ability to communicate that vision. So, essentially with AI, all designers have the opportunity to become creative directors.
As I’ve said before, will greatly change the graphic designer as we know her today. Your job is to evolve with AI to become the designer of the future, lest you be left behind in your career field.
Fave trick or technique for working with AI that you’d never want to be without again?
Sketch-to-Image AI. While most image generators started as text to image tools, many have evolved to reference sketches as well. This puts even more control in the hands of the designer, placing a premium on the skill of illustration. By referencing a sketch combined with a quality prompt, it’s far easier to achieve your vision. Stability AI (formerly Stable Diffusion) has tried to differentiate itself by prioritizing sketch to image features. That said, the same can be done in Midjourney by uploading a sketch to your prompt and using it as an image reference (clicking the image icon on the Midjourney web platform or using the image reference call in Discord).
Best suggestion or tip for anyone new to working with AI?
Don’t let inconsistent outputs get you down. AI is a fickle beast that is learning every day. If a prompt works one day yet returns completely different results the next, that’s par for the course. It’s up to you to experiment with different ways of prompting the machine to get your intended output. You might find yourself generating over 100 images just to get to one you like, but I promise it will get easier over time. The more you practice the more you can anticipate ways to tame the beast. So be patient, don’t get down on yourself, and—as usual with any new skill—practice makes perfect!
Most surprising and/or delightful way AI has been useful recently?
Agentic AI. We are on the precipice of a global revolution in how we work and interact with our daily tasks (among other things). AI agents like Anthropic’s Claude, Microsoft’s CoPilot, and Google Gemini Gems are a game changer and only the tip of the iceberg of what’s possible.
Consider this: designers must rely on content to inform their design, which means content can often be the bottleneck in any design process. But what if you could train an AI agent on your brand language, history, industry, and any other facts to make it a subject matter expert? Think of what that expert can do for you—research infographics, write spec docs, write ebooks, script motion graphics, the list goes on. Suddenly your bottleneck has been shattered and you get to be the hero—providing clients and stakeholders with more solutions and efficiencies than ever before.
This scenario isn’t just a dream… it’s already here! Hop onto Gemini Advanced to see how easy it is to create a brand expert Gem to see what I mean.
What are you most excited to share in your upcoming session?
I’m so excited about sharing what’s possible with AI driven video. In fact, if 2024 was the year of text to image AI’s rapid evolution, 2025 will be the year of image to video and text to video AI.
This article was last modified on September 22, 2025
This article was first published on February 21, 2025
Commenting is easier and faster when you're logged in!
