The Library of Congress this week drew attention to a collection of 19th-century watercolor architectural drawings, noting that such hand-crafted renderings “aren’t as common anymore.”
In a July 22 post on X, the institution shared examples from its holdings that depict the Library itself, the Biltmore Estate, and the White House.
🎨 Watercolor architectural drawings aren't as common anymore, but the Library of Congress collections contain many dating back to a time when they were, including some of the Library itself, the Biltmore Estate and even the White House. pic.twitter.com/OMDuWWwbYf
— Library of Congress (@librarycongress) July 22, 2026
Watercolor was long used to differentiate materials, create depth through shadow, and preview finished buildings or interiors. However, digital tools and generative artificial intelligence systems have largely replaced these methods for producing architectural visualizations.
Historical Watercolor Renderings
Watercolor has been applied to architectural drawings for centuries. According to the Library of Congress blog “Architectural Watercolor Works of the 19th Century,” the medium allowed architects to show color-coded materials and three-dimensional effects.
Examples include an 1807 ink-and-watercolor drawing by Benjamin H. Latrobe of the principal floor plan and site of the President’s House (White House), in which watercolor creates shadows for depth and incorporates landscaping. An 1817 Latrobe plan of the U.S. Capitol uses red shades to distinguish rebuilt walls and columns after fire damage from new construction, with gray indicating proposed elements.
Richard Morris Hunt produced a circa-1890 perspective rendering of the Banquet Hall at the Biltmore Estate in Asheville, North Carolina, and an 1891 dome section of the Administration Building for the World’s Columbian Exposition in Chicago.
Other works in the collections include a circa-1897 rendering of the Thomas Jefferson Building’s Northwest Corridor by G. W. Peters and a 1937 interior perspective of the John Adams Building reading room by J. Rowland Snyder.
Evolution Through Digital Tools
Hand-drawn and painted renderings remained standard into the mid-20th century. Computer-aided design (CAD) systems began entering widespread architectural use in the 1980s.
Autodesk released AutoCAD in December 1982, according to historical accounts of the software’s development, and subsequent programs such as ArchiCAD (1987) and later solid-modeling tools enabled precise digital drafting and early three-dimensional models.
By the 1990s and 2000s, specialized rendering software produced photorealistic computer-generated imagery, incorporating accurate lighting, materials, and camera perspectives. These tools accelerated production while reducing reliance on traditional watercolor and ink techniques.
Rise of Generative AI Imagery
Generative AI systems capable of producing images from text prompts entered broader public use in the early 2020s. OpenAI introduced DALL-E in January 2021 and released DALL-E 2 in 2022. Midjourney entered open beta in July 2022. These tools generate architectural visualizations, paintings, and other imagery in seconds.
Earlier computer-generated art predates these systems. Artist Harold Cohen developed AARON, described by the Whitney Museum of American Art as “the earliest artificial intelligence (AI) program for artmaking,” beginning in the late 1960s at the University of California, San Diego, and naming it in the early 1970s. Cohen continued refining AARON until his death in 2016. The Whitney presented an exhibition of the work from February to May 2024.
Artists’ Perspectives on AI-Generated Work
Some creators and companies that serve artists have stated positions on generative AI. Procreate, the developer of a widely used digital illustration app for iPad, published a statement titled “AI is not our future” on its website.
The company wrote that “Generative AI is ripping the humanity out of things. Built on a foundation of theft, the technology is steering us toward a barren future.” It added, “We’re here for the humans. We’re not chasing a technology that is a moral threat to our greatest jewel: human creativity,” and “Creativity is made, not generated.”
Procreate has stated it will not add generative AI features.
Coverage in outlets such as Creative Bloq has documented broader concerns among illustrators about training data scraped from existing artworks without licenses, as well as statements from organizations such as Fairly Trained opposing unlicensed use of creative works.
Other reports note that some practitioners use AI for rapid ideation or to reduce repetitive tasks while retaining human direction over composition and final output.
Ongoing Debate Over Definition
No single legal or institutional definition settles whether an AI-generated painting qualifies as art.
Traditional criteria have emphasized human intention, skill, and authorship. Generative systems produce images by statistical patterns learned from large datasets of existing works.
The practical shift from watercolor and drafting tables to CAD, photorealistic rendering, and generative AI tools is well documented across architectural and design practices. Questions of authorship, training data, and economic impact continue to be actively discussed among artists and organizations.
The Persistence of Hand-Rendered Work
Some architects and illustrators continue to produce hand-rendered watercolors and sketches for conceptual or presentation work, though these remain a minority practice.
Hand drawing has not disappeared. A 2024–2025 exhibition at the New York School of Interior Design, titled “The Persistence of Hand Drawing: Interior Rendering Today,” featured contemporary hand-drawn interior renderings by 12 New York-based architects and designers. Curators Donald Albrecht and Thomas Mellins noted that while computer imagery is ubiquitous, a number of practitioners still draw by hand, valuing its ability to enhance observation, convey scale and proportion, and create expressive representations of space, color, and light.
Artists such as Thomas Wells Schaller, an architect and watercolorist who has received awards including the 2024 Arthur Ross Award in Fine Arts from the Institute of Classical Architecture & Art, continue to produce and teach architectural watercolor.
Courses in traditional Beaux-Arts watercolor wash rendering are still offered by organizations such as the Institute of Classical Architecture & Art.
Current trends show digital and AI-assisted tools dominating commercial architectural visualization for their speed, photorealism, and ease of revision. Hand rendering is primarily used in high-end residential or classical design practices, early conceptual stages, client presentations seeking a distinctive or emotional quality, and by individual artists for personal or fine artwork.
Industry reporting indicates no widespread return to hand methods as the primary production tool. Instead, hybrid workflows are more common: practitioners may sketch or watercolor by hand for ideation and then refine or present digitally or use digital tools that simulate traditional media.
Pure hand-rendered architectural illustration is expected to remain a specialized, niche skill rather than a mainstream practice.
Provided by Dallas Express
