Taste framed as a competitive edge
An essay on the 'taste economy' argues that as generative tools make creation cheap, the ability to exercise taste—curation, editing and restraint—becomes a key differentiator for creatives. The piece positions taste as the skill that separates undifferentiated output from curated, repeatable design systems. (thevccorner.com)
A new essay argues that when artificial intelligence makes making things cheap, taste becomes the scarce skill: choosing what to keep, cut and repeat. (thevccorner.com) Ruben Dominguez made that case in an April 2026 essay for The VC Corner, framing “taste” as a business moat rather than a soft creative instinct. He wrote that generative tools now produce drafts, images and layouts fast enough that raw output is no longer the main constraint. (thevccorner.com) The shift sits on top of a real change in creative software. OpenAI said its GPT-Image models are built for “professional-grade” image generation and editing in ChatGPT and the application programming interface, while Adobe now sells Firefly as an all-in-one studio for generating and editing images, video, audio and design assets. (openai.com) (adobe.com) When more people can generate polished first drafts, the bottleneck moves from production to selection. Dominguez’s argument is that the advantage shifts to people who can set a visual standard, reject mediocre variations and turn one-off outputs into a consistent system. (thevccorner.com) That logic tracks how the tools are marketed. OpenAI’s developer documentation emphasizes controllable image workflows and editing, and Adobe describes Firefly as a way to accelerate ideation, exploration and production inside existing creative workflows rather than replace them outright. (developers.openai.com) (helpx.adobe.com) In practice, “taste” in this argument means curation, restraint and consistency. A creator can ask a model for 50 logo directions in minutes, but still has to decide which two fit the brand, which details to remove and which choices can be repeated across a website, ad campaign or product line. (openai.com) (adobe.com) The essay pushes back on the idea that more output automatically means more value. If everyone can generate endless options, undifferentiated volume starts to look like stock inventory, and the differentiator becomes a recognizable point of view applied over time. (thevccorner.com) That claim also has a limit. The same generative platforms that lower production costs can absorb parts of curation too, because they increasingly promise better control, more accurate text rendering and tighter editing loops inside the tools themselves. (openai.com) (adobe.com) So the essay’s bet is narrower than “humans stay essential.” It is that in a market flooded with competent drafts, the people who can impose standards — and build repeatable design systems from cheap generation — keep an edge that software alone has not standardized yet. (thevccorner.com)
