The floor for AI-generated website design is rising fast, and venture capitalist Ben Blumenrose is watching closely. As co-founder and managing partner at Designer Fund and a practicing designer for two decades, Blumenrose can still spot the difference between a site designed by a professional human and one produced by a model. The tells include an overabundance of "pills," floating dashboards, and gradients. But the gap is closing quickly. AI results today are "five times better than what the same kind of tools did a year ago," he says, and he expects the trend to continue. "Probably in six months to a year it's going to be very hard even for me to tell."
The rapid improvement has shifted the challenge in AI development. A few years ago, the industry focused on reducing obvious errors, such as ensuring AI-generated human hands had the correct number of fingers or that AI essays didn't mix metaphors. Now that basic correctness has become table stakes, the path to creative work that convinces or impresses an expert is less clear-cut. Taste, or at least an approximation of it, has become one of the next frontiers in model development. A host of startups are promising they can encode taste into AI, while major frontier labs are thinking about how to edge their models into subjective domains like writing and design.
Anthropic's Claude Design, which launched in April, has already encoded a particular aesthetic that is replicating across the internet. But as every human knows, even within a narrow profession, opinions on what is considered "good" can vary based on each person's mix of social, cultural, and work experiences. Developing taste in any creative field is both a technical challenge and a personal project. You need to learn enough information about what works from experts and from history, but then you need to develop an understanding of what you like and why. The same is true when teaching a model taste.
The first step in that process, information gathering and evaluating, is similar to how tech companies have operated for decades and how frontier labs have been developing model intelligence in recent years. The critical next step, deciding what feels good and unique and then making a choice based on discernment, is more foreign to the processes and systems for building technology. Still, technologists across the industry are attempting to tackle that qualitative quandary. At large and small scales, they are teaching models what "good" looks like for specific creative fields, aiming to hone AI taste to a fine-enough point. As it turns out, those methods rely on the taste of a whole lot of human professionals.
The first hurdle toward developing AI taste is getting the right data sets. Most general-purpose model providers like OpenAI and Anthropic have built their model knowledge bases by scraping the internet and gathering publicly available information. For a long time, the big labs have worked with "post-training," the process of refining an existing model to perform well at certain tasks. With the goal of developing generalized model intelligence, that training has been directed toward a best average. In the visual domain, this has resulted in reliable, if not terribly creative, output. But domain-specific expert-level taste requires domain-specific expert-level data, and a crop of established companies and new startups are looking to move past a generalist knowledge base to solve this problem in the creative space.
Figma, which uses foundation models in-house for productivity and production tools, notes that off-the-shelf models often have shortcomings. "Off-the-shelf models often lack the ability to judge design quality in a reliable way, and the amount of reasoning and resulting latency doesn't always correlate with the complexity of the task," says Figma's AI Research Lead, Sumithra Bhakthavatsalam. Frontier models can work too hard and for too long to produce less-than-tasteful results, so the company post-trains for faster outputs and less overthinking during iterative design work. It also tunes the model for overall quality, variety, and a deeper understanding of Figma and Figma's customers' design systems. To power that post-training, Bhakthavatsalam's team collects curated "best-in-class" examples of "great" and "good" from Figma's designers.
At Krea, co-founder Diego Rodriguez knows that general inputs won't yield results that his designer, filmmaker, and architect customers can use for their professional work. Krea offers a suite of tools for creative professionals, including image and video generators, and the output quality must be high. For Krea's visual generators, Rodriguez's team has gathered millions of references of film angles and styles, printing techniques and materials, and architectural materials and elements for post-training to ensure that the models they build with have the visual vocabulary for professional creative work. Architect Martin Frank Petersen used Krea to create early-stage visualizations of a multi-functional space in different scenarios, including a clay workshop and other uses, demonstrating the practical application of these trained models in professional settings.