How to Implement AI Workflows that Your Artists Will Actually Want to Use
AI is a powerful tool that can support and amplify an artist's creativity when deployed strategically and thoughtfully across an organization.
June 20, 2024
Introducing AI into creative workflows, particularly in game development, can feel like a double-edged sword. On one side, it offers unprecedented efficiency and new possibilities. On the other hand, it raises fears about losing control of the final product and compromising artistic integrity. The key is to implement AI in a way that both respects and enhances the artist's role, rather than replacing it.
Despite what you might hear from the tech industry hyping generative AI, there isn’t strong evidence that AI can replace the human creative process anytime soon. Because they can’t. No matter how sophisticated AI systems get, the nuance, intuition, and lifetime of human experiences create emotional depth that human artists bring to their work. That humanity is why art can make you feel things (and AI doesn’t feel!) So instead of a substitute for us humans, AI is a powerful tool that can support and amplify an artist's creativity when deployed strategically and thoughtfully across an organization.
AI for Creatives Means Generative Control
Most visual artists don’t think in the keywords and phrases that common generative AI tools call for; they more often think in visuals, emotions, and narratives. Asking an artist to use generic text prompts to generate entire images can feel unnatural and restrictive. While it can be useful in getting inspiration, tools limited to text prompts very often lose professionals due to a lack of effective interpretation and controllability. Ever yelled at a computer? That’s how it can feel for these artists!
Instead of pushing text-based chatbots onto artists, studios should focus on integrating AI tools that allow for high degrees of control and consistency. Tools should feel like a natural extension of their existing creative workflow, not a cumbersome add-on. Or they just won’t be used.
There are two broad categories of AI tools: simple text generators and more complex, customizable models and interfaces. Text generators can be great for early-stage brainstorming and inspiration, but they fall short when it comes to providing the control needed for later stages of development. Open-source models that can be customized offer more flexibility and ownership, aligning better with an artist's need for control. Further, the cutting-edge research in controllability is happening on open-source models, making them the first and typically the most capable in achieving the types of control professionals are looking to leverage.
Integrate Workflows, Don’t Replace Creativity
When evaluating how to incorporate AI into your team’s workflows, there are three key areas to evaluate and focus on solving for:
1. Spot Opportunities for AI Assistance
Begin by mapping your team’s existing workflow and identifying specific points in the creative process where AI can genuinely speed up work without compromising the creative process for your artists. Ask these artists what they find most tedious about their current workflow and you’ll immediately have a use-case for AI, and do so with confidence that your team will actually use it. For example, using AI to take an initial color rough and polish it using generative iteration into a more detailed rendering and compositing can save time without sacrificing artistic input and control.
2. Create Fluid Processes
Design workflows that allow for a seamless back-and-forth between AI outputs and human refinement. Most artists won’t use AI-generated content as-is. They want tools that help generate elements they can then integrate into their work, maintaining control over the final product.
3. Develop Custom AI Models
Custom AI models can be tailored to specific projects or styles, offering higher degrees of control. These models can be trained to understand the particular design language and aesthetic of a project, making them much more useful to artists. For example, a custom model can generate assets in a specific art style or for a particular intellectual property, ensuring quality.
Address Ethical Concerns Around Using Your Artists’ Creations
Let’s not beat around the bush: ethical concerns in AI are top of mind throughout the industry, especially when dealing with the content that your artists create. Artists need to be told explicitly how their work will be used in the future. If you plan to use your artists’ work to train future models for your studio, that can raise job security fears, so those concerns will need to be addressed openly and transparently with your team before you deploy the AI.
Foster a Culture of Experimentation
Encourage artists to experiment with AI tools and explore new creative possibilities. The best uses for AI tools will likely be organically discovered by an artist on your team experimenting with the tools in their free time. Provide time and resources for artists to play with the tools without the pressure of immediate production deadlines or “ROI” on their time spent in the tool. We’ve heard creative teams having success implementing informal creative jams or innovation days where artists can experiment with AI tools and share what they’ve learned with other teammates.
Provide Adequate Training and Support
Ensure that artists receive proper training on how to use AI tools effectively. Offer workshops, tutorials, and ongoing support to help them understand the capabilities and limitations of the tools. A dedicated support team, or an appointed expert user that can promptly address any issues or questions will accelerate the timeline to value. You could provide a series of video tutorials that walk through the AI tool’s features and best practices can help artists get up to speed quickly.
Highlight Success Stories
Showcase examples where AI has successfully enhanced the creative process of other artists. Many artists still view generative AI as a “push button, get picture” tool. Highlight examples of artists, particularly artists that your team respects, using generative AI as part of the creative process. For example, if an AI tool significantly sped up the character design process for a particular project of an artist in your studio, share that story with your team and have that artist demonstrate their use of the tool. Visually seeing an artist use the tool in a creative way can alleviate skepticism and encourage more artists to give AI tools a try.
Continuous Feedback and Iteration
Just like any other change management process, implementing AI workflows is an ongoing process that should be reviewed regularly. Schedule regular monthly or bi-weekly feedback sessions with your team, both in-person and anonymous, about what’s working and what’s not. Use this feedback to continuously improve the AI tools and workflows.
When your team feels heard and supported, they are more likely to embrace any new tool that significantly changes the way they work.
Conclusion
Integrating AI into creative workflows isn’t about imposing new technology on artists or minimizing their concerns. The most successful rollouts will acknowledge and respect these concerns, focusing on addressing them through a clear demonstration of value while preserving the artist’s creative control. Implementing AI workflows that truly resonate with your artists requires a transparent and compassionate approach that prioritizes artist’s needs and enhances their creative freedom.
By emphasizing collaboration, investing in custom solutions, ensuring robust security, and fostering a supportive and innovative culture, you can integrate AI in a way that empowers your artists and drives your studio’s creative success.
About the Author
You May Also Like