TT5: Two Lanes

May 7, 2026

This weekly email is my curated selection of interesting and useful topics related to BIM, design, and technology. I aim to provide you with some good information and a few laughs along the way. So, what did I find interesting this week? Read down to find out!

Yesterday I was a guest on BIM Pure Live for a deep dive into using Claude Code for Revit plugin development. If you've ever been curious about what vibe coding for Revit actually looks like in practice, the recording is a good place to start. It also ties into something I've been noticing in my feeds this week: the conversation about AI in AEC is splitting into two lanes. One lane is practical, asking what firms should actually be doing right now. The other is philosophical, asking what AI can and can't replace. This week's links cover both, plus a couple of pieces on the human side of getting better at your work.

Alright, here are five things to check out this week. 

#1: You Don't Have to Be AI-Native

KP Reddy makes the case that most AEC firms will never become "AI-native," and that's fine. What's not fine is doing nothing. He lays out five concrete moves any firm can make right now, including build a citizen coding program and train your team to think like owners.  It's a no-nonsense playbook that doesn't require a new business model or a Chief AI Officer.

Click here to read the full article at KP Reddy's Substack

#2: AI Won't Automate 80% of Architecture

Anthropic published a graphic showing that over 80% of architecture's tasks could theoretically be done by AI. This Architizer piece pushes back hard, arguing that the study confuses tasks with practice. The article draws on Kant, Heidegger, and a sharp quote from architect Chad Reineke to make the case that architecture requires taste, judgment, and civic responsibility, none of which can be modeled from a dataset.

Click here to read the full article at Architizer

#3: AI Terms Every AEC Pro Should Know

If you've been nodding along in meetings when someone says "agentic workflow" or "RAG," this one's for you. This is a plain-English glossary of the AI vocabulary showing up in AEC right now, organized into five sections: how AI works, how it takes action, where it gets its knowledge, daily-use concepts, and firm policy terms. Each term gets a definition, an AEC example, and a note on why it matters.

Click here to read the full article at ArchSmarter

#4: Learning That Actually Changes Performance 

Christopher Parsons interviews cognitive scientist Clark Quinn on the Smarter by Design podcast, and the conversation highights a problem most AEC firms share: learning programs that transfer information but don't actually change how people perform. Quinn argues that the ratio should be flipped to 20% content and 80% practice, and that firms need a full performance ecosystem rather than just more training.

Click here to listen to the episode at Knowledge Architecture

#5: How to Motivate Yourself to Do Anything

In this article, Scott Young breaks motivation problems into seven categories. The framework is useful because it stops you from treating every motivational stall as a willpower failure.  Sometimes the fix is making the work less painful. Sometimes the fix is admitting it's just not that important to you.

Click here to read the full article at Scott H Young

That's all from me. I hope you're having a great week. 

Michael