The retail apocalypse narrative is officially dead. U.S. retailers absorbed a net 5.5 million square feet in Q3 2025, according to CoStar, snapping up space faster than they're vacating it for the first time in years. Vacancy rates are at historic lows, new construction is tracking toward one of the weakest years in decades, and discount and grocery-anchored centers are leading the charge. Tight supply and sticky demand are making retail investable again.

The CRE recovery just hit a speed bump. October marked the first month of negative year-over-year transaction volume growth since early 2024, according to Moody's, as the buyer-seller stalemate drags on amid sticky rates and policy uncertainty. Still, institutional capital is circling discounts: New York Life scooped up a distressed Manhattan office building from BGO at nearly half its 2015 price.

The Fed cut rates 25 bps Wednesday, but the bigger story: restarting Treasury purchases with a $40 billion buy Friday and elevated purchases expected "for a few months." Call it stealth easing. For CRE, it's a mixed bag (near-term liquidity boost, but the dot plot signals fewer cuts ahead).

Chicago's industrial pipeline is roaring back with 11.3 million square feet under construction (up from just 1.5 million last year), putting it neck-and-neck with Atlanta. The catch: vacancy has climbed to 12.6%, one of the highest rates among major metros. Tenant demand remains active, but lease-up risk is real.

This might be the least sexy thing I've ever written about AI.

This one's going to sound pessimistic, but it's not. It's just honest.

Look around. People fall into two camps when it comes to AI:

Camp #1: The Doomers. They're worried. AI is coming for their jobs. Skynet is loading. They're one headline away from building a bunker and stockpiling canned goods.

Camp #2: The Utopians. These are the entrepreneurs. The optimists. They see AI as the golden ticket. Automate everything, outsource the grind, and coast into some new era of harmony. The spectrum here is wide. On one end, you've got guys dreaming of becoming a "single-CEO company" with 1,000 bots working 24/7. On the other, you've got Elon Musk floating universal income while robots make our piña coladas and add value to civilization on our behalf.

I call BS on both.

Not because AI isn't transformational. But because both camps are missing what's actually happening.

Here's what I've learned from 12 months of going deep on AI and watching our team at IndustrialIQ.AI build some genuinely incredible tools.

Yesterday, I underwrote a 22,750 SF small bay deal in Waukegan, Illinois using an AI model we've been developing.

In 10 minutes, I had a full OPEX analysis, rent roll validation, and a side-by-side comparison of the seller's numbers versus how the property would actually perform in our portfolio.

The seller's OM was light by $23,323 in operating expenses (a 15% NOI haircut they conveniently forgot to mention). 

Leasing costs: omitted. 

Grounds maintenance: omitted.

 Property management: understated by nearly $5,000.

Before AI, this analysis would take a couple hours. Maybe half a day on a complex one.

 Now? Ten minutes.

So here's where the utopians would say, "Great! Use that saved time to relax. Life is easier now."

That's not what happened.

I didn't kick back. I went deeper.

Doing things the old way, I’d send my assumptions to an underwriter, wait for a response, go back and forth with clarifying questions, and maybe schedule a tour in a few days once I felt confident the numbers made sense.

This time I had clarity in 10 minutes. I scheduled the tour immediately. Then I processed another 30+ properties.

Before AI, you'd only underwrite a deal when a seller signaled they wanted to sell. It was too time-consuming to do otherwise. 

Now, you can run high-level underwriting on prospects who aren't even selling yet, and cherry-pick the best candidates to pitch before they hit the market.

The speed doesn't just help you move faster on active deals. It lets you create deals that didn't exist before.

Here's the thing nobody tells you about AI: it's not "set it and forget it."

Working with AI is like collaborating with a savant. It's brilliant in a narrow domain, but it needs constant feedback. 

You're guiding it, correcting it, feeding it your assumptions, refining the output until it matches your thesis.

The more connected you stay to the flow of information, the better the answers get. The moment you disconnect, the output drifts.

This speed advantage isn't just yours. Everyone gets access to the same tools. Your competitors are 10x-ing their underwriting speed too.

So if you build a process that's 10x faster, 10x more accurate, and 1/100th the cost: congratulations. That edge only lasts about six months. Then it's table stakes.

The winners aren't the ones who automate and walk away. The winners are the ones who stay plugged into the information exchange, constantly learning, constantly iterating, constantly connected.

Here's how I see the next few years playing out:

Everything gets faster. Underwriting, sourcing, due diligence, leasing. All of it accelerates.

The painful stuff gets automated first. Data entry, comps, document review. The grunt work disappears.

You'll do more of what you're actually good at. Strategy, relationships, judgment. The stuff that can't be automated.

Information flow becomes the new skill. Your ability to stay connected (to process, interpret, and act on fast-moving data) is the edge.

The old game rewarded the person who could grind the longest.

The new game rewards the person who can stay in flow the longest.

AI won't make your life easier. It'll make it faster.

And faster means more reps, more decisions, more output. Not fewer.

If you're waiting for the robot piña colada retirement, I've got bad news. 

The game just sped up. And the players who stay connected to the flow are the ones who'll win.

That's my $0.02.