Oftentimes in the creative process, the most difficult part comes when transitioning from broad ideation to concrete ideas. When brainstorming, I often think in abstract terms, especially when ideas aren’t yet “put on paper.” This leads to overlooking important details, which only emerge when you start to think through the entire process. Another challenge arises in group settings, where coming to a shared understanding of a project can present a significant challenge before the work even begins.
This is where GenAI in the form of vibe coding tools truly can shine. While plenty of advice preaches caution against overreliance on these tools in production, their value in the early stages of prototyping is undeniable. GenAI helps make the journey from vague concepts to tangible outputs incredibly fast, helping me to visualize ideas quickly. For example, AI-powered design tools can code basic mockup apps in a fraction of the time it would take me manually, even with tools like Figma. With this it’s easy to explore (and discard) ideas with minimal time waste. Additionally, there now is a common reference point which team members can see and interact with.
Although on a more cautionary note, overreliance on the tool can probably degrade creativity. Especially when using it for ideation, the LLMs reliance on its training data will lead to “your idea” being fairly generic. Once in a class about future work a lecturer showed us a small scale experiment, where some students in her class were allowed to use ChatGPT and others weren’t for a creative writing task. While the quality of the texts where ChatGPT was used were consistently high, almost all the texts converged on similar themes, while the individually written texts were much more diverse. Still, there are ways around this (giving key words or themes, using the model with high heat) and the stories written with ChatGPT were well written afterall. So in conclusion, if used sensibly AI can really help expressing your creativity with tangible results.
