AI in the Trades: What's Real, What's Hype

A couple weeks ago I spent about forty minutes with members of the West Coast Lumber & Building Material Association talking about AI — session one of a three-part series we're doing with them. It's the kind of conversation I have constantly with distributors and manufacturers, so I wanted to put the core of it in writing too.

Quick background on me, for anyone who doesn't know: I grew up around the trades. My dad was in HVAC and plumbing and I've been renovating homes and running a small door-and-window distribution operation on the side since 2008. I've also spent most of my career on the technology side (Cisco Meraki, Verkada, Stripe) building the backend and frontend systems that move products from a warehouse to a customer's door. These days I run Matterhaul, but I showed up to that session as an educator, not a vendor, and I want to do the same thing here.

So, what's actually going on with AI in an industry like ours?

What AI is not

Before anything else, it's worth clearing out the noise. Three things I hear constantly that just aren't true:

• It's not a robot that replaces your team. Nobody's running a lumberyard with three people and an AI doing everything else. That pitch exists, and it's wrong.
• It's not a magic system that runs itself. AI needs context and direction. It still needs a human in the loop to be worth anything.
• It's not something only enterprise companies can afford. This one was true a few years ago. It isn't anymore. Twenty dollars a month gets you tools that would've required a real budget not long ago.

The one-sentence definition

Here's the whole thing, as plainly as I can put it:

AI is software that learns patterns from data and uses those patterns to make predictions or take actions. That's it. Everything else is marketing.

I mean that as a practical tool, not just a soundbite. Every time a vendor pitches you something, ask three questions: What patterns is it learning? What predictions is it making? What actions is it taking? If they can't answer those cleanly, that's your signal to walk away.

Not all AI is the same

I like to think about AI on a spectrum, because lumping it all together is where a lot of confusion comes from.

On one end you've got familiar AI — autocomplete, GPS navigation, spam filters. Things you've been using for years without thinking about it.

On the other end is overhyped AI — the "this will run your entire business" pitch. That's where a lot of vendors want to sell you, and it's mostly not real yet.

In the middle is useful AI — pricing-error checks, reorder signals, automated order-status updates, pipeline analysis. This is where I tell every operator to start. It's not flashy, but it's where the actual ROI lives today.

The most important distinction: this isn't traditional software

If there's one mental shift I want people to walk away with, it's this one.

Traditional software. Your ERP, your e-commerce platform, most of what you've bought over the years — is designed once and then fixed. Every business is different, but traditional software is built as one-size-fits-all, and changing it after the fact means a slow, expensive development cycle. It's a prefab house: it comes as-is, and adding an outlet later costs you real money and real time.

AI-native software works differently. It learns from your data and gets shaped around how you actually operate: Your workflows, your terminology, your exceptions. It's closer to a contractor who builds to your specs on the same foundation, and keeps adjusting as you go. That doesn't mean nine months and a six-figure budget to make a change. It can mean days.

This is the shift that actually matters for how you think about buying or building anything going forward — traditional software makes you conform to it; AI-native software conforms to you.

Where this leaves you

None of this requires you to become a technical expert. It just means you can walk into any AI conversation, with a vendor, with your own team, with your board, and ask sharper questions instead of reacting to hype in either direction.

In the next post, I'll get into the practical side: the questions to ask before you buy anything, and the red flags that tell you to walk away. And in session two of the LBMA series (July 22), we're going deep with a live member panel on where AI is actually paying off in distribution today — real use cases, not hypotheticals.

Got AI questions of your own? If you want to talk through what any of this looks like for your distribution or manufacturing business, reach out. I'm happy to talk it through.

— Shawn