AI in Design: Acceleration or Loss of Critical Thinking?

AI and Critical Thinking

In the design world, the use of AI language models (LLMs) has become common for quickly solving design problems. A designer recently noticed that he tended to rely too easily on these tools to generate high-fidelity designs in record time, a task he would have previously approached with deeper analysis. This realization prompted him to develop an application aimed at reclaiming and enhancing his critical thinking skills.

Critical Thinking in the Age of AI

Current AI tools allow for the design and creation of products at an unprecedented speed, but this rapidity can come at a cost. Clients and stakeholders often demand faster processes, but this does not necessarily guarantee better solutions. Sometimes, solutions are tested that should never have passed the initial verification stages. The testing and learning process is sometimes reduced to an approach of "throwing spaghetti against the wall to see what sticks," as the creation of these "spaghettis" has become inexpensive. As a result, critical thinking skills erode, as more and more reflection is outsourced to AI. Ironically, these skills will become crucial once the AI hype subsides, as they enable informed decision-making, asking tough questions, and evaluating beyond mere prompt construction.

Keeping Human Ideation

For this designer, the whiteboarding and sketching phase is essential in the design process. It is at this moment that he can quickly generate ideas and refine them. He feels most effective when he can exchange ideas with others who can critically question them and transform them into more developed concepts. However, often working remotely and on solo projects, he does not always have access to collaborators to help him build these ideas. This is why he considered leveraging AI to support this space for deep reflection, without compromising it. He began developing an AI collaborator capable of engaging in Socratic dialogue while sketching ideas, much like an excellent colleague would.

My Collaborator, My Terms

The creation of this AI collaborator application needed to meet very specific needs, those of the designer himself. It was crucial that the application operated under his conditions and supported his design process. This involved the application being able to see what he was drawing, understand the context, and engage with him in real-time. The goal was to allow ideas to flow freely, as they would in a workshop or meeting with a colleague. Uploading a sketch to an LLM and discussing it was not sufficient, as it resembled a cascading process, too slow and jerky, preventing a state of flow. The designer also wanted constructive criticism, not feedback that boiled down to automatic approvals. He aimed to eliminate weak ideas and build on the strongest ones, bringing in different perspectives to push and challenge them until they became worthy of further development.

Bringing Thia to Life

For his AI collaborator to be effective, it needed to work on design concepts with the designer, providing constructive criticism rather than merely generating ideas en masse. The designer chose to use Google AI Studio to build his application, a development environment that provided access to the real-time models he needed. This opportunity also allowed him to delve deeper into building AI tools beyond a simple Proof of Concept (POC) or a ghost product for user testing, into something he could actually use. After numerous prompt improvements, the designer brought his AI collaborator, named Thia, to life.

Sketching with Thia to Design New Features

The designer began using this first version of Thia to design Thia herself. He simply directed the webcam at his whiteboard, started sketching the ideas he wanted to explore, and engaged in Socratic dialogue with Thia. At this stage, he discovered that he began using adjacent AI tools for specific needs, such as Figma and Stitch, to explore design systems and animations before reintegrating them back into AI Studio. This process proved liberating and addictive, although it was easy to get sidetracked.

When Speed Becomes the Enemy of Reflection

The designer acknowledges that he would not have been able to build such a sophisticated application in just three days without the help of AI. The capabilities offered by these tools are truly immense. He was able to experiment and design features like animating Thia's breathing rhythm, a task he would have previously considered a mere "nice-to-have." However, with all the noise surrounding AI tools and the collective assumption pushing everyone to rush, the rigor of the design process is often neglected in favor of rapid construction. But what happens if you don't take the time to prove the concept from the start and simply build the wrong thing? Your competitors will surpass you with a superior product, regardless of whether you were first to market.

Critical Thinking Requires Friction

Stepping back, the real change brought by AI in product development is the speed at which we can execute each step. This does not mean we should stop seeking users to inform our solutions or reduce the time spent creating and refining robust ideas. It simply means we can accelerate the design process without skipping steps. The only way to maintain critical thinking skills is to use them, keeping them sharp — this means embracing cognitive friction and not just rushing to the finish line. Tools will become easier to use, but to build robust systems, we will need designers who understand how to ensure that products and value propositions are solid and support a healthy business model. This is where we need critical thinking skills to analyze, evaluate, and cross-pollinate ideas across different domains to create something greater than the sum of its parts. The cost of losing these skills is too high. For the designer, it was about keeping that sketching and ideation time sacred. The act of realizing an idea through physical drawing and writing is paramount; it literally builds memories compared to typing, activating parts of our brain that are essential for creativity.