Data centers are quietly rewriting the rules of construction, but not everything you've heard about them is true.

In this episode of The Punch List, hosts Jon Wright, TK, and Kanav Hasija bust the biggest myth about data center water use, break down why smaller AI models trained on proprietary data might beat Claude and OpenAI at real-world tasks, and unpack what HUD's new AI grant means for the future of plan review.

Along the way, the crew talks CapEx forecasts, the historical parallel between AI and the steam engine, and why AI might be the first technology that actually meets construction professionals where they already are, instead of forcing them to change how they work.

Highlights:
(00:00) Introduction
(01:15) AI spend shifts from proprietary to open source models
(02:45) The Bridgewater and Thinking Machines experiments
(06:53) Vertical intelligence vs general intelligence
(11:14) Data center CapEx forecasts and construction demand
(14:23) AI as a general purpose technology, lessons from the steam engine
(18:49) Busting the data center water use myth
(27:43) HUD's grant for AI-powered plan review
(31:36) Why AI fits construction better than SaaS ever did

Connect with us:

• Connect with Tanmaya Kala on LinkedIn: https://www.linkedin.com/in/tanmaya-kala-pe-3324a13/
• Connect with Jon Wright on LinkedIn: https://www.linkedin.com/in/jonpwright/
• Connect with Kanav Hasija on LinkedIn: https://www.linkedin.com/in/kanavhasija/

Resources:

• Coinbase CEO on AI spending cut by half at his org: https://x.com/brian_armstrong/status/2070670644577280109
• Custom trained models in enterprise brings more accuracy at way lower costs: https://thinkingmachines.ai/news/learning-to-replicate-expert-judgment-in-financial-tasks/

KP Reddy co-wrote the book on BIM for construction. Fifteen years later, he says BIM is dead.

In this episode of The Punch List, hosts Jon Wright, TK, and Kanav Hasija get him to explain why. They dig into how BIM turned into an expensive version of SketchUp, why empathy and understanding quietly decide whether a project succeeds, and what KP is building with ZeroRFI to pull construction back toward the "master builder" model.

The back half turns to policy and AI: the newly passed 21st Century ROAD to Housing Act, why manufactured and factory-built housing might finally have its moment, and what cheaper Chinese models like Kimi K3 mean for the future of data center construction.

TIMESTAMPS
(00:00) Introduction
(02:26) KP Reddy on the shift from "BIM is amazing" to "BIM is dead"
(08:20) AI slop, accountability, and the accuracy gap in construction data
(12:17) The two things every successful project needs
(20:25) Using LLMs to translate between owners, architects, and trades
(23:07) Building the master builder with ZeroRFI
(33:55) Inside the 21st Century Road to Housing Act
(41:45) Why factory-built and manufactured housing are having a moment
(45:22) Chinese open models, token economics, and data center demand
(51:25) Closing thoughts

Subscribe to The Punch List for weekly takes on AEC news, AI, and the future of the built environment.

Connect with us:

• KP Reddy on LinkedIn: https://www.linkedin.com/in/kpreddy
• KP Reddy’s website: https://kpreddy.co/
• Tanmaya Kala on LinkedIn: https://www.linkedin.com/in/tanmaya-kala-pe-3324a13/
• Jon Wright on LinkedIn: https://www.linkedin.com/in/jonpwright/
• Kanav Hasija on LinkedIn: https://www.linkedin.com/in/kanavhasija/

Resources:

• Biggest Housing Bill in 30 Years Becomes Law — Without Trump's Signature: https://www.npr.org/2026/07/10/nx-s1-5885027/housing-bill-without-trump-signature | https://www.cnbc.com/2026/07/11/21st-century-road-to-housing-act-homebuyers-sellers.html

In Episode 3 of The Punch List, Jon Wright, TK, and Kanav Hasija discuss Berkshire Hathaway's acquisition of Taylor Morrison, why OpenAI and Anthropic are preparing for public markets, and Autodesk's $3.6 billion acquisition of MaintainX.

They explore what Berkshire's move says about housing demand, whether AI companies going public will accelerate infrastructure spending, and why Autodesk may be trying to own the entire building lifecycle.

The episode also dives into Claude 4.8, token exhaustion, and practical ways teams can get more out of AI tools without paying more.

(00:00) Intro

(01:00) Berkshire Hathaway buys Taylor Morrison

(05:00) Why America still needs more housing

(08:00) OpenAI and Anthropic prepare for IPOs

(12:00) Will AI companies change their messaging?

(15:00) Autodesk acquires MaintainX

(18:00) Why Autodesk paid $3.6 billion

(20:00) Claude 4.8 and token limits

(23:00) Opus vs Sonnet explained

(26:00) How to stop running out of tokens

Connect with us:

• Connect with Tanmaya Kala on LinkedIn: https://www.linkedin.com/in/tanmaya-kala-pe-3324a13/
• Connect with Jon Wright on LinkedIn: https://www.linkedin.com/in/jonpwright/
• Connect with Kanav Hasija on LinkedIn: https://www.linkedin.com/in/kanavhasija/

Resources:

• Autodesk to acquire MaintainX, advancing unified platform in operations: https://adsknews.autodesk.com/en/news/autodesk-to-acquire-maintainx-advancing-unified-platform-in-operations/
• $960 Billion: Anthropic Beats OpenAI to the IPO Filing in a Race That Could Reshape AI Investing: https://finance.yahoo.com/markets/stocks/articles/960-billion-anthropic-beats-openai-161011329.html
• Berkshire's bet on Taylor Morrison suggests the housing market may have bottomed: https://www.cnbc.com/2026/06/01/berkshire-taylor-morrison-bet-housing-market-bottom.html