Daily field briefing
POLR AIPOLR AI

AI + AEC Today

Daily AI intelligence for the built world

Weekly digest

Weekly AI in AEC stories

← Archive

Edition · Monday, August 31, 2026

AI & Construction

Weekly Pulse · Industry actionability

The field is piloting, not just experimenting.

60%
ExperimentalPilotingField-ready
  • Physical AI moves onto the job site, from Caterpillar machines to design tools
  • AI agents start doing the paperwork, like reading and answering RFIs
  • The AI behind the tools keeps getting faster and facing new legal fights

Practical Use of the Day

Draft a clear RFI in two minutes

A good RFI gets you a fast, clear answer. A vague one bounces back and stalls the work. Let AI clean up your draft. Tell it the problem, the plan sheet or spec, and what you need to know. Ask for one short, clear question the design team can answer. Read it over, add your file numbers, then send. You get a sharp RFI without the blank-page struggle.

Try this prompt

You are a construction project engineer. Write a clear, short RFI. Problem or conflict: [describe]. Drawing or spec reference: [sheet or section]. What I need to know: [question]. Keep it professional and under 150 words. Ask one clear question, note any impact on schedule or cost, and do not invent details I did not give you.

Tool — Claude (or any capable LLM)

Stories

5 curated
ConTechHigh impact

Caterpillar puts its mining AI to work on more job sites

Caterpillar builds the big machines you see on job sites. For years it has run driverless trucks in mines. Now it wants to bring that same AI to more work. It made a Cat AI Assistant that techs talk to by voice. It helps them fix machines and find the right parts. Caterpillar plans to spend 100 million dollars to teach its workers AI and robotics. One operator may soon watch many machines from a single room.

Why it matters to GCs

Skilled operators are hard to find, and fewer people want to run a machine all day. AI that runs or watches machines could help fill that gap. It may also cut fuel use and keep crews out of harm's way.

TechCrunch·Aug 30·Read source →

LinkedIn
ConTechModerate

PlanRadar's new AI agents answer RFIs on their own

PlanRadar makes software to track job site work and papers. It just added AI agents that handle routine tasks for you. One agent reads a new RFI and searches your project files. Then it writes a reply and shows where each answer came from. Work that took hours can now take minutes. You can test an agent first and set what it may do on its own. Every action is logged, so you can see who did what.

Why it matters to GCs

RFIs pile up fast and slow the whole job. An AI that drafts the answer frees your team for real work. You still check it, so you keep control while saving time.

AEC Magazine·Aug 20·Read source →

LinkedIn
ConTechModerate

New AI tool turns a sketch or a few words into a 3D model

A startup called Illoca made a web tool named Plamo. You give it a sketch, a photo, or a short note. It builds a 3D model you can edit, not just a flat picture. The shapes stay smart, so you can change sizes and parts with ease. It can even add structure and MEP systems. The team calls it vibe modeling, like telling AI what you want. A free beta is out now.

Why it matters to GCs

Early design work eats a lot of time. A tool that builds a real 3D model fast could speed up options and bids. It gives your team a quick way to test an idea before you commit.

AEC Magazine·Aug 21·Read source →

LinkedIn
AI FrontierModerate

Music giants sue Anthropic over how it trained its AI

Sony Music and Warner Chappell sued Anthropic, the maker of Claude AI. They say Anthropic used their songs and lyrics without paying. The suit claims the AI was trained on copied, pirated work. Anthropic says it did nothing wrong and will fight the case. Earlier this year Anthropic paid 1.5 billion dollars to settle a book case. A court said training on real work is fine, but stealing it is not.

Why it matters to GCs

AI tools learn from huge piles of data, and the law is still catching up. These cases could change what AI tools cost and what they are allowed to use. It is smart to know where your AI vendor got its data before you lean on it.

TechCrunch·Aug 29·Read source →

LinkedIn
AI FrontierModerate

Anthropic shows AI that can improve its own training

Anthropic ran a test where AI helped make AI better. The system read past research, tried new ideas, and checked the results. It did this on its own, like a small research team. On 10 safety tests, the AI beat what skilled humans came up with. It did it in about six hours, for far less money. Anthropic says the model did not get worse as it improved.

Why it matters to GCs

AI is starting to speed up its own growth, so tools may get better fast. Better tools could reach the job site sooner than you expect. It is worth watching so you are not caught off guard.

TechCrunch·Aug 28·Read source →

LinkedIn