BRICKS & BYTES BULLETIN
INTELLIGENCE FOR CONSTRUCTION LEADERS

THIS WEEK
75% Skipped the Groundwork. Now AI Has Nothing to Work With

David Gal of Samsara says 75% of construction hasn’t started on AI. Plus: a third of global executives can’t find their own AI bill, and the data center boom hits a power wall.

THE EXECUTIVE BRIEFING
THIS WEEK’S KEY TAKEAWAYS

Full episode write-up at the bottom↓

Key Takeaway 1:

David Gal (Samsara) says roughly three quarters of construction hasn’t even reached the stage before AI. His four-stage model runs digitize, then insight, then automate. Skip stage one and the AI agent in stage three has nothing to work with.

Key Takeaway 2:

KPMG surveyed 2,145 executives across 20 countries: 29% have no idea where their rising AI costs come from. MIT found 95% of corporate AI projects fail on woolly goals and weak integration, the same groundwork gap David describes in construction.

Key Takeaway 3:

The data center boom keeps order books full, but transformer lead times have stretched to five years. Turner and Townsend say 87% of markets are short of specialist trades, the same trades every construction pipeline is fighting over.

7 THINGS WORTH YOUR ATTENTION
ON THE RADAR THIS WEEK

  • Esri’s User Conference opens Monday in San Diego, drawing 18,000 GIS professionals into BIM and digital-twin sessions. (More)

  • US CPI for June lands Tuesday, the first inflation read since the energy-price spike hit materials and labor costs. (More)

  • Rio’s inaugural DCPI-LATAM summit opens Tuesday, tackling the power and grid bottlenecks behind Latin America’s data center build-out. (More)

  • BuiltWorlds’ Construction Tech Conference runs Wednesday in Chicago, reading honest AI adoption across the jobsite. (More)

  • Siemens Transform opens Wednesday in Manchester, putting industrial AI to work across building and energy infrastructure. (More)

  • RICS’ India Real Estate Investment Summit lands Wednesday in Mumbai, tackling digital infrastructure and commercial real estate strategy. (More)

  • IEEE’s Power & Energy Society General Meeting opens Sunday in Montréal, powering the digital era through grid modernization. (More)

POWERED BY:

The multiplayer planning platform where construction teams plan together, stay aligned, and deliver projects faster.

The construction intelligence platform delivering data and insights to act faster and outperform, from site to boardroom.

The #1 construction management software for growing companies. Manage your projects from Tender to Handover.

FULL EXECUTIVE BRIEFING
75% Skipped the Groundwork. Now AI Has Nothing to Work With

An equipment manager at a large contractor spends this morning working a phone and a spreadsheet, hunting answers that sound simple everywhere else in the business. Where’s the excavator that was meant to be on the Leeds job? His company owns thousands of pieces of kit, and on any given morning, a good chunk of it goes unaccounted for.

Meanwhile, the same company’s board is asking how fast it can get AI into the business, because every board is asking that right now. Hold onto that gap, because it’s the thread running through this week’s episode with Samsara’s David Gal, and it explains why the AI reckoning landing on every other industry this year was never going to catch construction off guard.

Stage Zero Is Where Most of Us Still Are

We went deep on this one last Wednesday, so a quick recap. David puts roughly three quarters of construction at stage zero, running on clipboards and spreadsheets with no digital record of what they own. His four-stage model runs digitize, then insight, then automate, and skip the groundwork and the AI agent everyone wants has nothing to work with. The full breakdown is here, including the $16 million-plus in documented savings.

DEWALT’s own global research put a number on that gap this week: 90% of construction professionals believe AI will be indispensable within five years, yet only 8% use it day to day.

Read: The $16 Million Wake-Up Call Hiding in Your Equipment Yard

The People Signing the AI Checks Don’t Know What They’re Paying For

KPMG’s own numbers show the same blind spot spreads well beyond construction. A survey of 2,145 senior executives across 20 countries found that 29% had no idea where their rising AI costs were coming from, and a further third admitted their own confusion over pricing was getting in the way of using the tools well.

Usage-based pricing means the bill scales with how hard staff lean on the tools, and if nobody’s watching, the number at month’s end lands as a genuine shock. KPMG’s own quarterly research puts real-time cost visibility at well under a third of organizations.

MIT researchers found that roughly 95% of corporate AI projects failed to show measurable impact on the P&L, sunk by woolly goals and budgets chasing the easy win. The groundwork underneath the AI is what broke, the same groundwork David’s been talking about all episode.

The Data Center Boom Is Hitting Construction’s Oldest Bottleneck

Your order book leans on the data center boom staying strong, and the four biggest tech companies are spending north of $650 billion this year alone, real and committed money. But the build is starting to slip on the thing construction actually knows best.

Data Center Watch found at least 75 projects worth about $130 billion were blocked or delayed in the first quarter alone, the sharpest total the group has tracked since 2023, and New York has paused big data center permits for a year as local opposition keeps building.

A quieter bottleneck sits behind those headlines: transformers, switchgear, the electrical guts of the building, the power infrastructure now holding up projects that already have their chips and funding secured. Lead times have stretched from roughly two years before 2020 to as long as five now, so a developer can have the land and the money locked down and still be waiting on a transformer queue five years deep. To be fair, plenty of what’s cancelled was speculative from the start, and projects with land and kit locked down are largely still moving.

Even a firm that never touches a data center feels the squeeze. Turner and Townsend found 87% of markets short of specialist mechanical and electrical trades, and UK construction PMI logged its sharpest civil engineering fall in six years. Both are pulling on the same shallow pool of electricians and commissioning crews. Turner Construction’s own data center work drove a 44% revenue jump in one recent stretch even as the broader market stayed flat.

Where This Leaves Us

Every industry is running the same experiment this year, pouring money into AI and hoping the return shows up before the bill does. Construction just happens to be running it with real machines, real transformers, and a paper trail that either exists or doesn’t, and there’s no shortcut through that gap.

Pick one number to go find this week: what you spent on small tools last year, who owns the AI bill once it lands, or how exposed your pipeline actually is to a transformer queue that’s now five years long. Whichever one stings the most is the one worth starting with.

Reply and tell us which number you couldn’t find.


Powered by beehiiv