BLOG: AI Innovation at AU—What It Means for PICCO
- Dustin South
- Oct 2
- 3 min read
October 2025 – By Dustin South and Eva Wu
Dustin and Eva returned from Autodesk University in Tennessee with exciting insights into the future of design and make. This year’s conference spotlighted the growing role of AI in manufacturing, consulting, and building workflows, with a product focus on Autodesk Forma and Fusion.
Dustin connected with Autodesk Inventor professionals and developers, learned about upcoming product improvements, and discovered new ACC workflows. Meanwhile Eva explored opportunities with Autodesk and third-party developers on Autodesk Platform Services, with a strong focus on AI integration. Together, they gathered forward-looking perspectives on how AI may soon shape the industry in the coming years.
While these innovations are still in the testing and validation phase, examples of potential AI applications include:
Real-time wind and solar analysis to support sustainable design
Automated patterning of building elements to save time and reduce errors
AI assistants in Revit that can modify models, streamline RFIs, and accelerate documentation
Administrative AI tools for managing personnel and project access in Autodesk Construction Cloud
Data-driven AI for summarizing, synthesizing, and visualizing project information in new ways
AI linking with Microsoft programs to create reports, RFI’s and presentations
It was encouraging to see the level of support and community engagement that the technical and development teams provide to the industry through this event. Feedback on day-to-day workflows, new functions, and improvements was actively encouraged and even mature products like inventor are still being improved upon and supported.
Another exciting development is Autodesk’s potential introduction of their own Model Context Protocol (MCP), a connector that integrates AI applications with Revit models, ACC data, and local project files. With Autodesk’s custom MCP, PICCO sees the potential to build AI models trained on our specific workflows and industry data, which could help us work faster, and with fewer errors.
Autodesk is also aiming to further connect the building and manufacturing industries, opening the door for more seamless collaboration across the full design and construction lifecycle. PICCO will continue exploring these emerging technologies to evaluate their readiness and potential for adoption, ensuring we remain at the forefront of innovation for our clients and partners.
Opinion Piece
The recent emphasis on AI at Autodesk University left one major question unaddressed: cost. Popular AI chatbots are marketed at relatively low subscription fees, yet the economics remain unclear. Training large models and maintaining the infrastructure required to run them is enormously resource-intensive. The long-term sustainability of current pricing models remains uncertain. Whether AI will eventually become more efficient, consuming fewer energy resources and less data, is unknown.
It became clear that AI use cases were limited to information manipulation, preliminary design and computation. While actual daily modeling and drafting with AI were not yet possible or even practical. Many AI workflows showcased at the conference highlighted tasks already achievable through existing scripts. Wind and solar analysis, for example, was presented as an AI capability, despite longstanding availability through traditional tools. Renaming sheets to match a specific format is another case where a simple script provides a more efficient solution than an AI model that must process entire datasets to retrieve context. While natural language tools can lower the bar for users to create tools and customizations in their workflows, it seems that the use cases of AI that was showcased raises the questions: if computational costs outweigh the productivity gains, the economic justification falls apart.
The keynote attempted to reassure attendees with the assertion that “AI is not in a hype cycle.” The need for such a statement, however, suggests otherwise. The potential for AI in design and construction is undeniable, but its practical and economic value remains unproven, particularly when simpler and more affordable tools already exist.
Written by: Dustin South, Project Manager and Eva Wu, Lead Software Developer