Pat Alesandrini - Small Jurisdictions, Dark Store Theory, and Technology in Assessment

When Pat Alesandrini talks about artificial intelligence in assessment, he doesn't mince words: "You have to treat these programs like somebody who has absolutely no understanding of what we do for a living."

This isn't technophobia speaking, it's wisdom earned from decades in the trenches of mass appraisal. As immediate past president of IAAO and now working in Hillsborough County, Florida, Alesandrini sees AI as both our industry's greatest opportunity and a tool that demands respect.

The Six-Month Babysitting Rule

The assessment profession has always been about managing complexity at scale. We've gone from hand calculations to spreadsheets, from paper maps to GIS. Now we're standing at the edge of another transformation, one that promises to handle our mundane tasks while we focus on the properties that truly need human expertise.

But here's what engineers in Telford, England, taught Alesandrini: plan on babysitting your AI systems for six months to a year. Why? Because these systems will solve problems whether they have the right data or not.

"Let's say the database only has one-story homes that have sold," Alesandrini explains. "You're trying to value two-story homes and the system's going to look and say, 'I don't have any two-story homes, but I have to value this two-story home.' And it will come up with a solution."

That solution might be completely wrong, but the AI will treat it as gospel truth going forward, unless someone's watching.

From Horror Stories to Best Practices

We've all heard about the attorney who submitted AI-generated legal briefs citing fictional cases. In assessment, the stakes are just as high. Every value we produce affects real people, real businesses, real tax bills.

This is why IAAO, under Alesandrini's leadership, tasked its standards and research committees with developing guidelines for AI use in mass appraisal. It's not about limiting the technology, it's about harnessing it responsibly.

"It's a tool that we use," he emphasizes. "Just like your tape measure, if you use it right, you get great results. If you use it wrong, you're going to get skewed information."

The Small Jurisdiction Challenge

While large counties debate AI implementation, smaller jurisdictions face a different reality. Alesandrini, who started his career in Tazewell County, Illinois, hasn't forgotten where he came from.

"Anybody can appraise with 75,000 transactions," he notes about Hillsborough County's data-rich environment. But what about the township with 15,000 parcels and two sales in five years? What about the jurisdiction that achieved a coefficient of dispersion of 253, "almost mathematically impossible," yet devastatingly real?

These offices, often understaffed and underpaid, can't wait for perfect AI solutions. They need tools now. That's why IAAO is developing programs to help abstract cost per square foot, land values, and depreciation, not a full CAMA system, but practical tools for practical problems.

The Human-Machine Partnership

The fear that AI will replace assessors misses the point entirely. As Alesandrini sees it, AI frees us from the repetitive to focus on the complex. It's the difference between cranking through routine residential valuations and tackling that unique industrial property that keeps you up at night.

He recalls teaching multiple regression in the 1980s when almost no one had access to the computing power needed. "Now anybody who has a computer with Excel can do multiple regression in a product they've picked up for a thousand bucks."

The same democratization is happening with AI. The question isn't whether to use it, but how to use it well.

Dark Stores and Bright Futures

Even as we embrace new technology, old challenges persist. The dark store theory, where big-box retailers argue their occupied stores should be valued like abandoned ones with deed restrictions, still haunts assessment offices from Michigan to Florida.

"There's a reason that store moved," Alesandrini points out. Demographics shift, markets change. But the solution isn't legislation banning certain valuation theories. It's assessors who understand highest and best use, who can argue deed restrictions, who know their craft deeply enough to defend their values.

This is where AI can't help us, yet. These complex arguments require human judgment, market knowledge, and the ability to see beyond the algorithm.

Advice for the Next Generation

For those entering the profession now, Alesandrini's advice is timeless: education and experience. Take the core classes. Find mentors willing to invest in you. Stretch yourself, even if it means appraising cemeteries when you've never done it before.

"If you don't ask, you're never going to learn," he says, recalling how he partnered with experts, read every article he could find, and wasn't afraid to admit what he didn't know.

Most importantly, work somewhere that values your growth. "If I was working for a firm that didn't want to invest in me, I'd probably look somewhere else."

Key Takeaways

• AI requires active supervision: Plan on six months to a year of careful monitoring before trusting any AI system with autonomous decisions.
• Data quality drives AI quality: Systems will create solutions even with bad or missing data, they just won't be the right solutions.
• Small jurisdictions need different tools: Not everyone has 75,000 sales. IAAO is developing practical solutions for data-poor environments.
• The profession isn't going anywhere: AI handles the routine so assessors can focus on the complex properties and situations that require human judgment.
• Education never stops: Whether it's learning AI, defending against dark store appeals, or stretching into new property types, continuous learning defines professional success.

As Alesandrini notes, he may not be around to see AI's full impact on our profession, but his excitement is palpable. The tools are getting better, more accessible, more powerful. The question is whether we'll use them wisely.

The answer, as always in assessment, comes down to the professionals willing to do the work, to babysit the algorithms, defend the values, and never stop learning. The machines may be artificial, but the intelligence guiding them must be very, very real.