Architecture Firms Build Proprietary AI as Data Remains Locked Away
_fastcompany · reddit · 2026-08-27
Architecture may be one of AI's hardest pursuits due to data inaccessibility. Unlike coding or writing, core data for architecture models—digital drawings, 3D models, and sketches—is stored in private firm servers, not the open web.
- Current State: Major AI labs are not tackling this challenge, leaving it to the firms that hold the data.
- Corporate Action: Architecture firms are actively building AI capabilities by hiring data scientists, hosting internal AI idea competitions, and developing bespoke plugins/apps to automate tasks. Some are even creating hyper-niche LLMs to generate forms and floor plans reflecting their signature styles.
More from Companies & People
- Metriport raises $26M to fix healthcare's data problem, deployed at Amazon and GM — ilyasu · 2026-08-28
- Google deserves credit for deploying AI to consumers at scale for free — andreisavu · 2026-08-28
- Alex Rives, pioneer of protein language model ESM, named to TIME100 AI — proteinrosh · 2026-08-28
- Every Built KateBench: Cloning a Colleague With 30,000 Edit Data Points — every · 2026-08-28
- Cohere Chief AI Officer Joelle Pineau Named to TIME 100 AI — cohere · 2026-08-28
- Vinod Khosla asked about fusion misconceptions and commercialization timeline — PierceLilholt · 2026-08-27