Page 2
Architecture has its own AI tools, its own problems, and its own version of vendors promising things that don't work in production. The overlap with mechanical CAD AI is smaller than you'd think.
No. But it will change what the job looks like, which tasks feel tedious, and which skills keep you employed. Here's what actually matters.
Yes. And here's why that answer won't change for a long time.
Machine learning has been in CAD for longer than the marketing suggests. Feature recognition, mesh cleanup, and constraint solving all used ML before 'AI' became a line item in every vendor's pitch deck.
Vendors promise a lot. Research papers promise more. Here's what I think will actually ship, actually work, and actually matter in the next five years.
The vendor version: AI is transforming design. The reality: AI is automating some annoying tasks, generating simple geometry, and making search slightly less terrible. That's still useful. It's just not a revolution yet.
Training AI to generate CAD models requires CAD training data. Most of the good data is locked inside corporate vaults. What's publicly available is limited, biased, and often missing the metadata that matters most.
AI is faster at generating simple geometry. Humans are better at everything else. The interesting part is where the boundary actually sits right now.
A real design process has concept, detail design, DFM review, documentation, and revision. AI fits into maybe two of those stages. Here's which ones.
A year ago, everyone predicted AI would revolutionize CAD. Some predictions were right. Most were early. A few were just wrong. Here's the honest scorecard.
